Lung ultrasound-guided management of acute breathlessness during pregnancy

Authors


Correspondence to: L. Zieleskiewicz
Email: laurent.zieleskiewicz@ap-hm.fr

Summary

Lung ultrasonography is a standard tool in the intensive care unit and in emergency medicine, but has not been described in the particular setting of the labour ward. During pregnancy, acute respiratory failure and pulmonary oedema are not uncommon life-threatening events. We present two case reports outlining the potential of lung ultrasonography in parturients. In case 1, lung ultrasonography allowed early diagnosis and treatment of acute dyspnoea in a parturient admitted for suspected asthma exacerbation. Lung ultrasonography revealed a ‘B-pattern’ of vertical lines radiating into the lung tissue, indicating severe pulmonary oedema complicating previously undiagnosed pre-eclampsia. In case 2, a pre-eclamptic patient was managed with combined transthoracic echocardiography and lung ultrasonography. The accuracy of lung ultrasonography in detecting interstitial oedema at a pre-clinical stage allowed adequate fluid resuscitation in this patient who had a high risk of alveolar pulmonary oedema. We believe that these cases strongly support the prospective validation of lung ultrasound for management of lung disorders in pregnant women.

Acute dyspnoea during pregnancy may be encountered by obstetricians, critical care specialists and anaesthetists [1]. The diagnosis of dyspnoea can be a challenge and determines subsequent treatment. During pregnancy, the causes of dyspnoea include non-cardiogenic pulmonary oedema secondary to asthma, embolic disorders, pneumonia or exacerbation of underlying disease, or cardiogenic pulmonary oedema [2]. Lung ultrasonography is a standard tool for the diagnosis of pulmonary disorders [3, 4] and acute pulmonary oedema [5]. As lung ultrasound does not use ionising radiation it is safe to use during pregnancy, but to our knowledge has not been reported in this setting. We report the use of lung ultrasound in two pregnant women with acute respiratory failure.

Case number 1

A 40-year-old pregnant woman, gravida 3 para 2, was admitted to our hospital at 31 weeks’ gestation in acute respiratory distress. She had asthma, morbid obesity (pre-gestational body mass index 41 kg.m−2), and diet-controlled gestational diabetes. Her previous pregnancies were uncomplicated. Before her admission, she had suffered from breathlessness for one week, which her general practitioner considered to be an acute exacerbation of her asthma.

On examination, she was tachypnoeic with an arterial oxygen saturation of 85% breathing room air. Auscultation of the lungs revealed wheezing with bilateral basal crepitations. She had bilateral lower limb oedema, but a normal neurological examination. Her arterial pressure was 180/90 mmHg and heart rate was 120 beats.min−1. Prednisolone 1 mg.kg−1 and salbutamol via nebulisers were started immediately. Blood tests showed a haemoglobin concentration of 8.7 g.dl−1 and albumin of 29 g.l−1. Renal and liver function tests and platelet count were within the normal range.

Urinalysis was positive for protein. Proteinuria of 8 g.day−1 was confirmed later on 24-h urine collection. Fetal ultrasound showed intrauterine growth retardation with biparietal diameter and femur length less than the fifth percentile. Fetal monitoring showed a baseline heart rate of 140 beats.min−1 with normal variability. A diagnosis of severe pre-eclampsia was made. The patient remained under the joint care of the anaesthetic and obstetric teams.

Lung ultrasound was performed to determine the cause of respiratory failure. A Philips CX 50 compactXtreme ultrasound system (Philips Medical Systems, Suresnes, France) with an abdominal curved probe (C5-1 with pure wave technology, Philips Medical Systems, Suresnes, France) was used. With the parturient lying in the supine position, the probe was positioned in a longitudinal orientation perpendicular to the ribs. Multiple abnormal B-lines, vertical lines radiating into the lung tissue, were found in four different anterior rib spaces on each side (Fig. 1). This pattern indicated significant pulmonary oedema [6]. A quantitative estimation of the severity of the condition was carried out by counting the number of B-lines when scanning at the parasternal, mid-clavicular, anterior-axillary and mid-axillary lines from the second to the fourth intercostal space on the left side and the second to the fifth intercostal space on the right side (28-rib space technique) [7]. The count was 118, confirming a high level of extravascular lung water and the severity of the pulmonary oedema.

Figure 1.

 B-pattern on lung ultrasound is indicated by multiple anterior confluent B-lines (arrows). This indicates severe pulmonary oedema.

Lung ultrasound was followed by transthoracic echocardiography (TTE). Left ventricular ejection fraction was assessed at 60%. Left ventricular filling pressures were calculated using Doppler imaging. The E/e′ ratio (peak early mitral valve inflow velocity divided by early diastolic velocity at the mitral annulus) was high at 11.6 (normal value < 9.5 during pregnancy) [8]. Right-sided structures were normal. The combined ‘heart and lung’ evaluation allowed us to go through the differential diagnosis of acute dyspnoea. In this case, acute pulmonary oedema, rather than exacerbation of asthma, of asthma, was the cause of her respiratory failure.

The management included oxygen therapy 5 l.min−1, administration of furosemide 20 mg intravenously every 8 h and nicardipine 4 mg.h−1 intravenous infusion for hypertension. Clinical signs (visual disturbances, headache, hyperactive tendon reflexes, epigastric pain) and blood markers (haemolysis, liver enzymes and platelet count) of pre-eclampsia were regularly checked, and these remained unchanged throughout her hospital stay. She was given steroid treatment with betamethasone to promote fetal pulmonary maturation and an insulin infusion was started.

Her respiratory status showed rapid improvement after diuretic administration and fluid restriction. Repeat lung ultrasound showed a change to a normal pattern with appearance of horizontal A-lines and disappearance of the B-lines (Fig. 2). Repeat TTE showed normal left ventricular filling pressure.

Figure 2.

 Normal lung ultrasound scan after diuretics. Pleural line (black star), rib shadow (white star), horizontal A-lines (arrows).

Oxygen administration was reduced to 3 l.min−1. The stabilisation of her condition allowed delay for 48 h before caesarean section. This was performed using spinal anaesthesia. Respiratory symptoms resolved during the first postpartum days and oxygen therapy was tailed off over 48 h. She was discharged from hospital with oral nicardipine medication.

Case number 2

A 20-year-old primigravid woman at 35 weeks’ gestation was admitted to our institution with severe pre-eclampsia. On admission, her blood pressure was 167/101 mmHg and heart rate was 87 beats.min−1. History and clinical examination revealed flashing lights, a continuous headache and exaggerated tendon reflexes. Her urine analysis showed protein concentration of 3.7 g.l−1 with a normal urine output. Blood tests were normal. Lung ultrasound did not reveal any B-lines. A TTE showed normal filling pressure with an E/e′ ratio of 5.47. Intravenous nicardipine was administered and titrated up to 3 mg.h−1.

The obstetric team made a decision to induce labour. Two hours after admission, she developed eclampsia. Magnesium sulphate and clonazepam were started to treat seizures as she was transferred to the operating room for urgent caesarean section under general anaesthesia. This was induced with thiopental 5 mg.kg−1 and suxamethonium 1 mg.kg−1. She developed post-partum haemorrhage with a haemoglobin concentration of 6.8 g.dl−1, an International Normalised Ratio of 3.5, fibrinogen concentration of 0.19 μmol.l−1 and platelet count of 137 × 109. l−1. She was transfused six units of red blood cells, six units of fresh frozen plasma, fibrinogen 1.5 g and tranexamic acid 1.5 g. The control of haemorrhage required bilateral radiological embolisation of the uterine arteries.

At the end of the procedure, the patient was awakened. She complained of moderate dyspnoea, with an arterial oxygen saturation of 96% breathing air. Breath sounds were decreased and bilateral basal crepitations were still present. Her blood pressure was 142/81 mmHg and heart rate was 107 beats.min−1. An electrocardiogram showed sinus tachycardia.

Lung ultrasound showed multiple bilateral sliding B-lines (Fig. 3). The four anterior rib space images per side showed a B-pattern [6], that is three or more B-lines on each of the four anterior scans. The number of B-lines counted with the 28-rib space technique was 72 [7].

Figure 3.

 Pulmonary oedema of lesser severity than Fig. 1 is shown by multiple anterior nonconfluent B-lines (arrows).

Assessment using TTE showed a high E/e′ ratio of 9.75. This confirmed the diagnosis of pulmonary oedema with elevated filling pressure. Based on the clinical and ultrasound findings, her resuscitation included oxygen administration at 3 l.min−1 by non-rebreathing mask, and furosemide 20 mg.

After 2 h of treatment, the patient had voided 600 ml urine. Her respiratory status had considerably improved with clinical resolution of the pulmonary oedema. Repeat ultrasound investigations showed disappearance of the B-pattern and a decrease in the E/e′ ratio to 7.47. Her further progress was uneventful with no recurrence of pulmonary oedema or features of pre-eclampsia.

Discussion

In the emergency setting, lung ultrasound is now a standard tool for the diagnosis of acute dyspnoea [9]. It was initially described with micro-convex probes with low emission frequency [5]. However, abdominal and cardiac probes are also widely used and even high-frequency linear probes such as those used for regional anaesthesia can provide a good view [10].

For a lung ultrasound examination, the probe is placed perpendicular to the ribs. The posterior shadowing of the superior and inferior ribs is seen. With a normal pattern, the pleural line is identified as a horizontal hyperechoic sliding line just deep to the rib line. The lung sliding results from the movement of the parietal pleura against the visceral pleura during the respiratory cycle. Beyond this pleural line, horizontal artefacts called A-lines are normally seen. In the case of interstitial syndrome (usually pulmonary oedema), the increased density of lung tissue creates reverberation artefacts called B-lines. B-lines are defined as discrete laser-like vertical hyperechoic lines that start from the pleural line, extend to the bottom of the screen without fading, and move synchronously with lung sliding [3]. B-lines were initially known as comets. Pulmonary oedema may be rapidly diagnosed by scanning the anterior chest and identifying three or more B-lines on each side; this is called a B-pattern. A more detailed approach is taken in the 28-rib space technique. The number of B-lines is counted at four sites (parasternal, mid-clavicular, anterior-axillary and mid-axillary lines) in each space from the second to the fourth intercostal spaces of the left hemithorax and the second to the fifth intercostal spaces on the right. This allows a semi-quantitative assessment of severity of the interstitial syndrome.

Studies have shown that in adult dyspnoeic patients, lung ultrasound can easily differentiate between bronchial disorders with a normal pattern and acute pulmonary oedema with a B-pattern [11]. Pulmonary oedema is a leading cause of acute dyspnoea during pregnancy [2]. Lung ultrasound is more accurate than chest radiography in the diagnosis of pulmonary oedema [12]. It is of similar accuracy to computed tomography [13], but the lack of irradiation is a clear advantage in the obstetric setting. Our case studies illustrate that it may be a readily available tool in the early diagnosis of acute dyspnoea in pregnant patients. In the first case, lung ultrasound ruled out the initial diagnosis of asthma in favour of pulmonary oedema, allowing improved obstetric and anaesthetic management.

Pulmonary oedema in pregnancy is a life-threatening condition [14], which may have cardiogenic or non-cardiogenic aetiology [15]. Non-cardiogenic causes include pre-eclampsia, tocolysis, and intravenous fluid administration. Fluids must be administered carefully as pulmonary oedema is strongly associated with positive fluid balance [16, 17]. In our second case, pulmonary oedema may have been precipitated by delivery, fluid overload, red blood cell transfusion, magnesium sulphate or hypoalbuminemia [15]. Lung ultrasound showed rapid transformation from dry interlobular septa (normal pattern) to wet interlobular septa (B-pattern) concomitantly with the first clinical signs [5, 6]. Lung ultrasound could be a useful tool for preventing unnecessary and potentially harmful fluid expansion in patients at risk of pulmonary oedema since interstitial oedema, the silent phase preceding alveolar oedema, can be easily and quickly diagnosed [6].

Lung ultrasound evaluation can be performed at different stages of the peripartum period known to be at risk, such as before and after fluid expansion and during delivery, allowing early initiation of adequate treatment [18]. Moreover, lung ultrasound is closely correlated with extravascular lung water and wedge pressure [6, 19, 20].

Lung ultrasound is a non-invasive technique requiring no specific knowledge in echocardiography and there is a steep learning curve for diagnosis of interstitial syndrome [3]. One study showed that a brief training module of 1 h was enough to achieve good recognition of images of pulmonary oedema for physicians both with and without previous ultrasound experience [21]. Lung ultrasound and TTE are probably complementary in the assessment of the haemodynamic profile of unstable patients. Lung ultrasound can predict the risk of pulmonary oedema, but not fluid responsiveness. Conversely, TTE provides more information about circulating volume status, fluid responsiveness [22] and cardiac contractility, but does not provide information on pulmonary oedema. Furthermore, specific training and equipment are required for TTE. We believe that TTE and lung ultrasound should ideally be combined, but that the use of lung ultrasound on its own should be further evaluated to guiding fluid management in obstetric patients.

In conclusion, the learning points from these cases are that lung ultrasound: (i) can aid the rapid identification of the cause of acute dyspnoea in pregnant women; (ii) is an accurate tool for the assessment of pulmonary interstitial oedema; and (iii) may be a useful tool in the management of pre-eclamptic parturients. We believe that all anaesthetists should be aware of the use of lung ultrasound in the management of pregnant patients.

Acknowledgements

Published with the written consent of the patients.

Competing interests

No external funding and no competing interests declared.

Ancillary