ABSTRACT
Unexplained hypoxia in a pregnant patient is an alarming finding for patient and provider. The differential for hypoxia is broad, and many imaging techniques and procedures are contraindicated in pregnancy. Transient pulmonary arteriovenous malformations (AVMs) are a rare and relatively poorly studied cause of hypoxia in pregnancy. Our patient is a 27-year-old G1P0 female with a remote history of asthma who presented to clinic with slowly progressive exertional dyspnea, exertional tachycardia, and paroxysmal nocturnal dyspnea. She reported use of a home oximeter which read in the high 80s% during exertion. Prior to presentation, the patient was evaluated in the Emergency Department and noted to have an oxygen saturation of 86% on room air. A transthoracic echocardiogram, computed tomography angiography of chest, and basic laboratories including B-type natriuretic peptide, troponin, COVID-19, and hemoglobin were unremarkable. Her clinical timeline is presented in Figure 1. Further testing was obtained, including pulmonary function testing, bubble echocardiogram, nocturnal oximetry, and shunt study. Work-up revealed a 15-20% shunt, depending on calculation, and insignificant desaturations during nocturnal oximetry. Her symptoms progressed, and repeat nocturnal oximetry showed marked overnight desaturations requiring supplemental oxygen for the remainder of her pregnancy. She delivered a healthy baby girl via cesarean section without serious complication. Repeat testing in the post-partum period showed resolution of nocturnal desaturations and decreased shunt, and her exertional dyspnea and desaturations resolved spontaneously. This case illustrates the challenging diagnosis of transient pulmonary AVM in pregnancy. Case reports of this phenomenon have been published, but as in our case, no definitive diagnosis was made secondary to testing limitations in pregnancy and quick resolution of symptoms in the post-partum period. Some reports describe pre-existing pulmonary AVM becoming worse during pregnancy causing hemothorax, fetal demise and even death. While the mechanism is not known, theories include the vasodilatory effects of progesterone on vascular smooth muscle as well as other physiologic changes in pregnancy such as increased plasma volume. These AVM are thought to be like those seen in hepatopulmonary syndrome. Similar to our case, increasing positional hypoxia has been reported as the pregnancy progresses. Further research dedicated to early and accurate detection of pulmonary AVMs in pregnancy is necessary. (Figure Presented).