Seasonal Influenza A (H1N1) Infection in Early Pregnancy and Second Trimester Fetal Demise

A second trimester fetal demise followed influenza-like illness in early pregnancy. Influenza A virus (H1N1) was identified in maternal and fetal tissue, confirming transplacental passage. These findings suggested a causal relationship between early exposure and fetal demise. Management of future influenza outbreaks should include evaluation of products of conception associated with fetal loss.

A second trimester fetal demise followed infl uenza-like illness in early pregnancy. Infl uenza A virus (H1N1) was identifi ed in maternal and fetal tissue, confi rming transplacental passage. These fi ndings suggested a causal relationship between early exposure and fetal demise. Management of future infl uenza outbreaks should include evaluation of products of conception associated with fetal loss. I ncreased maternal illness and death appeared in early reports of infl uenza A pandemic (H1N1) 2009. Three pregnancy-related complications associated with pandemic (H1N1) 2009 were reported in May 2009: 1 postpartum maternal death, 1 preterm birth, and 1 early second trimester (weeks [14][15][16][17][18][19][20] loss (1). In April 2010, a summary of pandemic (H1N1) 2009 among 788 pregnant women demonstrated a disproportionately high risk for death; the rate of spontaneous miscarriage was 1.4%, but details were not provided (2). Correlating infl uenza exposure to pregnancy loss is not straightforward because fi rst trimester miscarriage is common, second trimester loss before 24 weeks is not well studied (3), and viral identifi cation in products of conception has rarely been attempted. We report a second trimester fetal demise that occurred after exposure to seasonal infl uenza A virus (H1N1) early in pregnancy.

The Case
In 2008, a 30-year-old primigravida physician experienced intrauterine fetal demise (IUFD) at 20 weeks' gestation, as established by last menstrual period and confi rmed by ultrasound at 8 weeks. The patient spent most of her pregnancy providing direct patient care. In October 2007, during gestational weeks 2-6, she was exposed to at least 2 patients with presumed infl uenza. She reported symptoms of fever, myalgia, headache, and cough during gestational week 4. She had no testing for infl uenza, did not receive an-tiviral therapy, and took acetaminophen; her symptoms resolved. She received the killed seasonal infl uenza vaccine at gestational week 6. An ultrasound at 18 weeks demonstrated early growth restriction and oligohydramnios with normal appearing fetal kidneys and head. IUFD was noted at 20 weeks.
The patient had an unremarkable medical history, did not smoke, and was normotensive throughout pregnancy. Antibodies to cytomegalovirus and toxoplasmosis were not detected.
The immature placenta showed no visible abnormalities. Microscopically, histiocytes were abundant in the maternal space (chronic intervillositis) and were noted within the fetal chorionic villi (Hofbauer cells; Figure 1  infl ammatory changes or viral inclusions were identifi ed in the placenta or fetal organs. Chronic intervillositis was evident in the placental histology and has been previously described with nonspecifi c changes in miscarriage and recurrent pregnancy loss (4). Formalin-fi xed placental tissue was imaged with electron microscopy. Histiocytes identifi ed from the maternal intervillous space and fetal chorionic villi demonstrated characteristics of viral production ( Figure 2). Several well-formed viral capsids were noted within the cytoplasm, each containing regularly spaced projections along the surface of the virion corresponding to the hemagglutinin and neuraminidase spikes.
PCR for M1 capsid protein confi rmed infl uenza A. Immunohistochemical testing was performed on tissue from the placenta and on a sampling of fetal organs by using infl uenza A virus (H1N1)-specifi c antibody. In the placenta, stippled immunofl uorescence ( Figure 1, panel B) was noted in the histiocytes of the intervillous space (maternal) and fetal intravillous histiocytes. In the fetal organs, immunofl uorescence for subtype H1N1 was notable in the surface epithelial cells and scattered luminal histiocytes of the fetal respiratory and gastrointestinal tracts. No immunofl uorescence was noted within the fetal organs, which implied transamnionic passage of virus (amnionic fl uid infection). The fetus was XY karyotype. Y-specifi c immunofl uorescent stain with histiocyte counterstaining was performed. Because only the intravillous (fetal) histiocytes stained positively, maternal and fetal infl ammatory responses were confi rmed.

Conclusions
Harris reported higher rates of miscarriage after exposure and infection in the earliest months of pregnancy in the Spanish infl uenza outbreak of 1918 (subtype H1N1) (5). Hardy et al. reported a similar observation following the Asian infl uenza outbreak of 1957-58 (6), as did investigators in the United Kingdom during the 1985-86 infl uenza season (7).
Several case reports document transplacental transfer of infl uenza virus. Yawn et al. documented infl uenza A transplacental transfer from the mother to amniotic fl uid and fetal heart (8). Mel'nikova et al. summarized placental fi ndings associated with infl uenza infections in publications from 1987 and 1994, describing perivillous and villous infl ammation with concurrent amniotic fl uid immunofl uorescence for infl uenza virus (9). In another study (10), researchers from Beijing, People's Republic of China, reported autopsy fi ndings for 2 patients (1 of whom was 4 months pregnant) who died of avian infl uenza A (H5N1). Viral antigens were present in the maternal pulmonary system, and virus was detected in placental Hofbauer cells (histiocytes), cytotrophoblast, fetal lung, circulating mono-nuclear cells, and liver macrophages (10). Their documentation of tropism of the infl uenza virus to placental tissue is similar to the histopathologic features in our report. Our unexpected discovery of viral production in placental, maternal, and fetal histiocytes is another example of transplacental passage of infl uenza.
Miscarriage or completion of pregnancy may follow nonfatal maternal infl uenza infection. Adverse fetal outcomes have reportedly included congenital malformation and schizophrenia, but factors that determine these effects remain unclear (11,12). Fetal consequences of infl uenza exposure may be an effect of systemic infl ammatory response to the infection, which could represent direct action of the virus on the placenta, the maternal-fetal interface, or the fetus. Uchide et al. reported the infl ammatory response of decidual and fetal tissue in the presence of infl uenza virus infection and demonstrated expression of mRNA for a set of proinfl ammatory cytokines such as interleukin-6, tumor necrosis factor-α, and granulocyte-macrophage colony-stimulating factor, secreted in substantial amounts in response to exposure to infl uenza virus in vitro (13). As our case suggests, a complex interaction of maternal and fetal responses at the interface is a more likely explanation. A total of 15%-20% of pregnancies end in spontaneous miscarriage, most before gestational week 12 (3). Approximately 50% of miscarriages are associated with chromosomal defects, which leaves many other miscarriages unexplained (14). Redline et al. examined fi rst trimester miscarriages not associated with chromosomal abnormality, observing histologic changes of chronic intervillositis and increased perivillous fi brin (15). The mechanism of loss was thought to be related to a maternal infl ammatory response to infectious antigen or an autoimmune phenomenon. Whether these fi ndings can be extrapolated to second trimester losses is unknown. Second trimester miscarriages account for only 1%-2% of all pregnancy losses (3). Histopathologic features of these early second trimester miscarriages are not well studied, and descriptions are usually included in reports of fetal losses during 14-27 weeks, which include IUFD (>20 weeks) and miscarriage. Consequently, a direct causal relationship between infl uenza A and fetal demise has not been well established.
In our report, a retrospective investigation of the patient's products of conception with ultrastructural, immunohistochemical, and molecular methods that identifi ed subtype H1N1 implies an exposure during her fi rst trimester. Type-specifi c serologic analysis was not possible because she received infl uenza vaccine several weeks after presumed exposure. The maternal and fetal infl ammatory responses (chronic intervillositis) after placental infection may have reduced villous surface area for oxygen and nutrient transport, resulting in growth restriction and ultimately second trimester loss.
Investigation of the effects of infl uenza in early pregnancy is needed to evaluate potential pathophysiological relationships between maternal exposure, infection, and fetal loss or adverse outcome. During the 2009-10 infl uenza pandemic, the Centers for Disease Control and Prevention (Atlanta, GA, USA) collected information from populations exposed to and infected with pandemic (H1N1) 2009 virus. Products of conception were not included among the samples cataloged. We recommend that future infl uenza outbreak evaluations of women exposed to infl uenza with fi rst or early second trimester pregnancy loss include collections of placental and fetal tissue samples.
Initial report data were presented at Frontiers of Pathology, September 2008, University of Michigan Department of Pathology, Ann Arbor, Michigan, USA.