Electrical impedance tomography effectively used in a case of paediatric pulmonary Langerhans cell histiocytosis

Abstract

A 2-year-old boy with severe pulmonary Langerhans cell histiocytosis presented in extreme respiratory failure. He was intubated and ventilated. Despite maximal support, he deteriorated and needed extremely high ventilator pressures. An electrical impedance tomography monitor was used to inform management. This is a monitoring technique which is not used in children due to the lack of suitable interface devices and a lack of randomised clinical evidence. Despite technical difficulties, a good signal was achieved. This informed management and enabled the selection of a suitable ventilator strategy, facilitating weaning. Electrical impedance tomography is a viable technology for use in paediatric critical respiratory failure. This is a non-invasive and safe technology which adds individual patient information which is not available through any other modalities. We urge equipment manufacturers to develop belts which will allow routine application of this life-saving technology in children.

Background

Pulmonary Langerhans cell histiocytosis (PLCH) is an extremely rare condition affecting the lungs of primarily adult smokers. It is characterised clinically by a cystic degradation of the lungs, which is often severe.¹ The prevalence in children is unknown, but is extremely low, probably under 1 in 10 million children.

With suitable chemotherapy, PLCH is a treatable condition with an overall good survival rate. However, the clinical course of the condition is one of severe lung degradation in the initial few months, leading to often unsustainable respiratory failure. Extracorporeal life support is not an option as the time scale to improvement is too long. Lung transplant is also not an option as the time scale of deterioration is too short.

The respiratory failure in a severe case of PLCH is extreme. Most of the lung tissue can be involved, leaving a negligible amount of functional alveoli available for gas exchange. Any deterioration over the basal state can be fatal.

Electrical impedance tomography (EIT) is a relatively established technology, comprising a non-invasive monitor which is primarily used for respiratory function analysis.^{2 3} It consists of an array of electrodes which circle the chest, where each electrode pair is used to sequentially apply a small AC current. The voltages between the remaining electrode pairs are detected and change with the impedance of the lungs. From the collected set of voltages, a two-dimensional picture of the chest, measuring 32×32 pixels, can be reconstructed and displayed in real time with a frame rate between 20 and 50 Hz. As the impedance of inflated lung differs to that of deflated lung, both the global volume curve and regional distribution of ventilation can be visualised.

So far, EIT has not been widely used clinically in children, except in research studies.⁴ Commercial belts of the correct size do not exist. However, there is no theoretical block to its use. Indeed, the nature of a child’s chest means that the signal is likely to be more accurate than the adult, as the distance between two electrodes is smaller, the cross-section captured by the belt is smaller, there is generally less adipose tissue and the ribs are thinner and more conductive.

Case presentation

A 2-year-old previously well boy presented with a spontaneous pneumothorax. His parents are non-smokers. His chest X-ray showed cystic changes. He recovered fully with a chest drain. He had a CT scan and a lung biopsy and was discharged home. He then represented after an episode of cardiac arrest following bilateral tension pneumothoraces. On the same day, the diagnosis of PLCH was made. After stabilisation, he then had a period of frequent pneumothoraces needing drainage, but remained at a relatively well baseline (figure 1).

Figure 1.

Coronal and axial CT lung at induction of chemotherapy.

He then developed pneumonia. He was intubated and ventilated. He rapidly became extremely difficult to ventilate. He did not tolerate high frequency oscillation. He was refused for extracorporeal membrane oxygenation and refused for urgent lung transplant as he was too unwell. He was eventually relatively stabilised on airway pressure release ventilation (APRV) mode.

However, over the following hours, he continued to deteriorate. His arterial oxygen saturations were in the 70s despite a fractional inspired oxygen (FiO₂) of 1.0. His oxygenation index was 46.9. Parents were informed of the bleak prognosis and prepared for his likely death. The organ donation team were involved.

Over the next 48 hours, he gradually stabilised and then slowly improved. His oxygen saturations crept into the 80s. He was still maximally ventilated with high pressure low release time APRV. Serial chest radiography excluded pneumothoraces.

The treating team then faced very difficult treatment decisions in a still highly unstable patient with a tiny margin for error. Although he had clinically improved, the fragility of his lungs meant that ventilator strategy was difficult. Making the wrong ventilator change would have been potentially disastrous.

At that time, we had access to an EIT device (Draeger Pulmovista 500) on compassionate grounds. However, this did not come with belts small enough for a 14 kg child. Instead, a series of 16 ECG electrodes was placed around the chest and the electrode belt attached to these. Parents were fully informed of the experimental and off-label nature of the procedure prior to use and were strongly supportive.

EIT data

Images were immediately available and informed ventilator strategy. Within hours of attaching the EIT monitor, significant progress had been made allowing a more physiological ventilator strategy.

The technical nature of attachment of the EIT monitor using ECG electrodes created some challenges. The child had bilateral chest drains and a Hickman line, meaning the chest already had multiple bandages and fixings. We were able to create a relatively unobscured line of electrodes around the chest. Due to the tensions on the monitoring wires, a bandage was placed around the chest to maintain contact. Skin integrity was closely monitored by the attending nurse (figure 2).

Figure 2.

Electrical impedance tomography in use in this patient.

Data was saved to the EIT machine and then available for later analysis using the Draeger EIT Data Analysis Tool 6.3. For the above images, both the dynamic and status images were set to a tolerance of ±8. The whiter the picture in each pixel, the more change there is in the impedance signal (equivalent to air movement). The absolute impedance number is equivalent to the total chest expansion.

At the first time point, a clear imbalance of ventilation is seen on the tidal image (figure 3). He was on APRV at 34/0 cm H₂O, with an FiO₂ of 0.80 and an inspiratory/expiratory (I:E) ratio of 3.4:1. Tidal volumes are relatively low (as seen by the relatively dull tidal image), but global expansion is very high. The breath-to-breath waveforms show a highly non-compliant lung, with little change in compliance with each breath. This is a very overdistended lung.

Figure 3.

Screenshots of electrical impedance tomography (EIT) data, showing at three difference time points: ventilatory settings, tidal image view, global EIT signal trend and breath-by-breath EIT waveforms for a shorter time period for the whole lung and each of four quadrants.

He was changed to biphasic positive airway pressure (BiPAP) assisted spontaneous breathing (ASB) mode, pressures 34/10 cm H₂O, FiO₂ 0.80, Ti 1.0 and I:E ratio 1:1. The second image shows improved balance of ventilation, with significantly reduced global expansion. Tidal volumes are higher, especially in the right posterior quadrant. Global expansion is variable over the time of the recording, but the individual breaths are much more normal, with good inflation and deflation seen.

He was further weaned to BiPAP ASB, 30/10 cm H₂O, FiO₂ 0.65, Ti 1.0 and I:E ratio 1:1.2.

The final image shows excellent (in context with the lung pathology) regional ventilation balance. Global expansion is at a steady state, with three of the four lung quadrants showing relatively normal looking breath shapes.

The clear progression of balanced lung ventilation is visible through time. A more balanced, regional ventilation is also visible, although some significant areas of hypoventilation remain.

Outcome and follow-up

Eighteen months after his episode of critical ventilation, the patient is at home and off oxygen. Externally, he seems like a normal 4-year-old boy. He is fully active and developing normally. Some cystic lesions remain in his lungs, and some scarring is found on CT scan (figure 4). He has completed his chemotherapy. His prognosis is excellent.

Figure 4.

Equivalent CT scan slices to figure 1 following successful treatment for pulmonary Langerhans cell histiocytosis.

Discussion

This 2-year-old boy had extraordinarily difficult lungs to ventilate. The natural history of paediatric PLCH is one of eventual resolution, given the correct chemotherapy. However, the challenge for intensivists is to maintain life while the lungs are at the nadir of their function.

As can be seen from the above CT scans, our patient had around 90% lung destruction. His respiratory reserve was extremely low, and when he developed pneumonia his functional alveoli were further degraded. Making any ventilatory change was very high risk, and conventional monitoring would necessarily have a significant lag time. Saturations would be dependent on the development of atelectasis, which will take time to develop. Blood gas analysis could realistically be done hourly. Chest radiography needs movement of the patient and could only be done 2–3 times per day maximally. Our fear was that this lag time may mean he would set on a clinical path of deterioration which would be irretrievable.

The treatment of any patient who is ventilated requires a deep understanding of respiratory physiology. The manipulation of the ventilatory pressures and volumes requires an appreciation of how the ventilator works, and of how the body would typically react to any changes. However, in cases of extreme ventilatory failure, where the lungs are highly unusual in their physiology, even expert physiologists may not be able to accurately predict how an individual’s gas exchange will react to changes in ventilatory strategy. In patients where individual areas of the lungs have different time constants, a change which may be beneficial to one area may be detrimental to the other. The only way to ascertain overall benefit is to try a change and then assess later.

The great advantage of using EIT in this patient was the instant breath-to-breath feedback after any ventilator change. We were able to change from being tentative to being bold. Each change was instantly and reactively monitored. Benefit in some areas could be offset by deterioration in others. The minute volume numbers from the ventilator will only be able to give a total ventilation volume, without appreciation of the regional effects, or the global expansion.

As with all patient monitoring techniques, EIT is not curative or therapeutic, but it can assist the clinician in making changes based on more patient knowledge. A full explanation of the use of EIT is beyond the scope of this case report, but in this case, we were able to use the EIT data in the following ways:

• Improvements in regional ventilation were instantly visible and gave positive feedback allowing continuation of strategies showing improvement and cessation of negative strategies.

• Improvements in respiratory system compliance were instantly visible.

• Global hyperinflation changes were visible, allowing moderation of pressures to improve the patient’s situation on the respiratory system compliance curve.

There were technical difficulties with this particular application of the device. Applying the ECG dots was complex, and required the patient to be rolled, with the inherent risks of such movements in a critically ill child. The patient’s skin could tolerate approximately 2–3 hours continuously, before a rest period was needed. Maintaining good skin contact was challenging. A bandage needed to be wrapped around the thorax to maintain electrode pressure.

Until the technicalities of a suitable connection system for attachment to a paediatric sized thorax is developed, this technology will not be generally adopted. However, we have demonstrated clear benefit to the treating team from the use of EIT, with real-time knowledge of how ventilatory changes affect lung physiology. There is no other modality of lung monitoring which gives a similar level of immediacy, accuracy and convenience.

Learning points.

• Electrical impedance tomography is a novel technology to clinicians which adds continuous information on ventilation which is not available through other means.

• It is a viable, useful and safe technology even in small children.

• Patients at the edge of viability with critical lungs are able to benefit from such monitors.