Favourable outcome after prolonged veno-venous extracorporeal membrane oxygenation (V-V ECMO) support for Pneumocystis jirovecii pneumonia in a renal transplant recipient

Abstract

A 27-year-old man, with a history of renal transplantation, presented with acute kidney failure and Pneumocystis jirovecii pneumonia. The patient developed severe acute respiratory failure and required support by veno-venous extracorporeal membrane oxygenation for a total of 59 days. During this period, the patient had extremely low tidal volumes using a lung protective ventilation strategy and intermittent prone positioning was used to optimise oxygenation. There was full recovery of pulmonary and partial recovery of renal function.

Background

Pneumocystis organisms are commonly present in the lungs of healthy individuals, where they are not associated with pneumonia. Pneumocystis jirovecii pneumonia (PJP) is mostly seen in immunocompromised patients and is thought to be the result of airborne spread from (healthy) humans, rather than from reactivation.¹ It is best known in patients with HIV, but is increasingly seen following chemotherapy or in recipients of stem cell or solid organ transplantations. The reported incidence in solid organ transplant recipients is 0.3%–2.6% prompting prophylactic treatment in the first year after renal transplantation.^{1 2} Respiratory effects of PJP are often more severe in HIV-negative patients, with more frequent occurrence of acute respiratory failure, but the duration of symptoms is often shorter (4–8 days).^1–3 The mortality rate of PJP in renal transplant recipients has decreased from 50% before 1990 to about 14% now, which is largely attributed to the widespread use of prophylaxis in the first year after transplantation, but can be as high as 80%–87% in patients requiring mechanical ventilation.^{1 2}

To the best of our knowledge, there is only one case report on the use of extracorporeal membrane oxygenation (ECMO) as a supportive strategy for a post-transplantation PJP.⁴ Most of the presented case reports focus on HIV-related PJP.^5–15 After a prolonged period of ECMO support, our patient showed full recovery of pulmonary and partial recovery of renal function, indicating that veno-venous ECMO (V-V ECMO) in combination with lung protective mechanical ventilation is a viable supportive strategy in the treatment of PJP.

Case presentation

A 27-year-old man, who received a living-related renal transplantation 4 months prior to presentation, was admitted to the hospital with acute renal insufficiency. Following the renal transplantation, the patient received immunosuppressive therapy (everolimus, tacrolimus and prednisolone). Antibiotic prophylaxis with trimethoprim/sulfamethoxazole had been discontinued shortly after the renal transplantation because of an increase in serum creatinine levels. The patient reported common cold with a productive cough, which started 9 days before admission.

Investigations

Serum creatinine increased from 181 μmol/L to 379 μmol/L. A biopsy showed no signs of acute rejection, but did show arteriolar hyalinosis, which can be a sign of drug toxicity. A chest X-ray showed bilateral atypical infiltrate and a CT scan revealed an extensive ground-glass pattern in all lung areas with bilateral basal consolidations, most prominent in the right lower lobe (figure 1). No bronchoalveolar lavage was performed prior to initiation of treatment, as the clinical condition of the patient was not considered good enough (marginal oxygenation). Following admission to the intensive care unit (ICU), a bronchoalveolar lavage showed a Pneumocystis and cytomegalovirus (CMV). A PCR for CMV in blood was negative. A rhinovirus was identified via a throat swab. The trough concentration of tacrolimus was within range (6.5 μg/L; target 5–7 μg/L), while the trough concentration of everolimus was above the required target (9.6 μg/L; target 6 μg/L).

Figure 1.

CT study on day of intensive care unit admission shows an extensive ground-glass pattern in all lung areas with bilateral basal consolidations, most prominent in the right lower lobe.

Differential diagnosis

Given the recent renal transplantation and the increase in serum creatinine, an acute rejection of the transplant was considered, but excluded by renal biopsy. Infectious diseases in the first months following renal transplant are often caused by opportunistic pathogens, of which PJP is the most prevalent. The findings on imaging studies, in combination with the discontinuation of trimethoprim/sulfamethoxazole prophylaxis and presence of a productive cough, prompted treatment for PJP prior to confirmation. Other opportunistic pathogens, such as CMV, Epstein-Barr virus (EBV) and Aspergillus fumigatus, were ruled out. The possibility of an everolimus-induced pneumonitis was considered, for which everolimus was stopped. Common symptoms of everolimus-induced pneumonitis include dyspnoea, cough, fatigue and fever. On imaging studies, focal areas of consolidation at the lung bases or ground-glass opacities are most often seen, although diffuse ground-glass attenuation or consolidative opacities can be observed.¹⁶

Treatment

Treatment for PJP was initiated using trimethoprim/sulfamethoxazole (2400 mg, once a day) and prednisolone (40 mg, two times per day). Due to a progressive hypoxaemia, the patient was admitted to the ICU for supportive therapy with high-flow nasal oxygenation (Optiflow).

Outcome and follow-up

Following a progressive respiratory insufficiency, the patient was intubated 3 days later (with a targeted tidal volume of 6–8 mL/kg ideal body weight). When the patient triggered his own breaths, large tidal volumes were seen in combination with signs of high breathing effort. In the following days, the requirements for respiratory support increased (driving pressure and oxygen demand). He was put on pressure-controlled ventilation with an inspiratory pressure of 32 cmH₂O and positive end-expiratory pressure of 10 cmH₂O, with a respiratory rate of 32/min. Tidal volumes were around 5.5 mL/kg ideal body weight and fraction of inspired oxygen fluctuated between 0.75 and 0.90. Oxygenation did not improve on continuous muscle relaxation or prone positioning. The ratio between oxygen pressure and fraction of inspired oxygen (pO₂/FiO₂) deteriorated from around 150 mmHg at the start of mechanical ventilation to below 90 mmHg, which prompted the initiation of V-V ECMO to allow for lung-protective ventilatory support. A femorojugular approach was used with a 29 Fr drainage cannula positioned in the left femoral vein and a 22 Fr return cannula positioned in the right jugular vein.

A progressive renal insufficiency required start of renal replacement therapy by continuous veno-venous haemofiltration. CMV load in serum remained undetectable, there was one short episode of EBV reactivation (with a maximum load of 3.2 log) and the second bronchoalveolar lavage showed a weak positive galactomannan test, for which treatment with liposomal amphotericin B was started, which was discontinued when the PCR was found to be negative.

With the mechanical ventilation on lung protective settings (positive end-expiratory pressure of 15 cmH₂O, driving pressure of 10 cmH₂O, respiratory rate of 10/min and fraction of inspired oxygen of 40%), tidal volumes were very low (figure 2) and the abnormalities seen on imaging studies were severe (figure 3). However, with the use of intermittent prone positioning, the patient received adequate respiratory support on V-V ECMO. Between day 10 and 16 after the start of V-V ECMO, an effort was made to wean ECMO support and recruit the lung. However, this resulted in a bilateral pneumothorax, which was attributed to barotrauma due to increased driving pressures. It was decided to return to lung protective ventilation and continue V-V ECMO for oxygenation and ventilation.

Figure 2.

Evolution of tidal volumes. V-V ECMO, veno-venous extracorporeal membrane oxygenation.

Figure 3.

CT study on day 39 of intensive care unit admission shows bilateral near-complete consolidation with air bronchograms, some small areas with ground-glass aspect and a small apical pneumothorax with subcutaneous emphysema.

After several weeks of very low tidal volumes and need for V-V ECMO, the possibilities of a lung transplant were explored. The renal insufficiency and the current critically ill status of the patient were considered relative contraindications for lung transplantation. Furthermore, the life expectancy after lung transplantation is limited, and, therefore, any viable conservative option would be preferential in this young patient. It was decided to continue the current support to see if the pulmonary status would be reversible.

After 7 weeks of V-V ECMO support and intermittent prone positioning, the tidal volumes started to increase under lung-protective ventilator settings. The patient was weaned from V-V ECMO and was decannulated on day 59. At this time, there was a slow return of renal function and renal replacement therapy could be reduced, although there was an intermittent need for haemodialysis for another week. Several days after decannulation, ventilatory support was discontinued, and 2 weeks later, the patient was transferred to the general ward.

The patient showed a continued improvement with a satisfactory return of pulmonary and renal function. The abnormalities on imaging studies were largely diminished (figure 4). He suffered from critical illness neuropathy, which improved following therapy. At the 1-year follow-up, the patient was living independently, had partially returned to work and was able to exercise. His kidney function was impaired, but stable (creatinine levels were around 350 μmol/L) and he is currently being screened for new renal transplantation.

Figure 4.

An X-ray during the recovery phase (2 weeks after intensive care unit discharge) shows improved air retention with a decrease of consolidative opacities.

Discussion

There is limited literature available on the use of V-V ECMO support for PJP, especially in renal transplant recipients.⁴ In order to perform V-V ECMO, there must be no severe left heart failure, as V-V ECMO does not directly support cardiac function. However, initiation of V-V ECMO will improve right ventricular failure by lowering pulmonary arterial pressures (through correction of hypoxaemia and hypercapnia) and reducing intrapleural pressures. There are different cannulation strategies to perform V-V ECMO; in most cases, two separate cannulas are used, as was done in our patient. Preferably, these cannulas are placed percutaneously using a Seldinger technique. A drainage cannula (for drainage of blood from the venous system to the ECMO system) is placed in the common femoral vein, and a return cannula (for return of oxygenated and decarboxylated blood to the right atrium) in the right internal jugular vein; this is known as the femorojugular approach. Alternative approaches include the use of two femoral cannulas (femorofemoral approach) or a double-lumen cannula in the right internal jugular vein.

As the ECMO system is used for oxygenation and decarboxylation, during V-V ECMO support, pulmonary gas exchange is not strictly necessary, which allows for (ultra-) lung-protective ventilator settings. However, a possible limitation of V-V ECMO is the occurrence of recirculation of oxygenated blood, the contribution of which increases as the ECMO flow increases.¹⁷

Long-term V-V ECMO support can be used to maintain adequate gas exchange for the duration of treatment of the underlying disease. This case shows that recovery of pulmonary function is possible, even after extensive lung damage.

The lungs of patients requiring ECMO may be particularly susceptible to ventilator-induced lung injury.¹⁸ The importance of lung-protective ventilatory strategies during ECMO support is further illustrated by a case where a spontaneously breathing patient generated very high transpulmonary pressures, despite ECMO.⁵

There is no definitive consensus on the optimal lung protective ventilation strategy during ECMO support. Guidelines from the Extracorporeal Life Support Organization (ELSO),¹⁹ recent reviews^{20 21} and the EOLIA trial²² recommend a ventilator strategy that limits alveolar strain by lowering tidal volumes (<4 mL/kg ideal body weight), reducing plateau inspiratory pressures (20–25 cmH₂O), using moderate amounts of positive end-expiratory pressure (10–15 cmH₂O) and reducing the fraction of inspired oxygen (0.30–0.50). Recommendations on respiratory rates differ between the ELSO guidelines¹⁹ and reviews^{20 21} (4–10 breaths/min) and the EOLIA trial²² (10–30 breaths/min); we aimed for a limited respiratory rate to minimise alveolar strain.

In this case, a restrictive driving pressure was used, which resulted in very low tidal volumes for a longer period of time. Increasing driving pressure in an effort to wean ECMO support was unsuccessful and possibly harmful. To improve oxygenation with protective ventilatory settings, intermittent prone positioning was used for recruitment and drainage of pulmonary secretions.^{23 24} No adverse events were seen with prone positioning.

The duration of V-V ECMO support for HIV-related PJP ranges from 4 to 69 days,^5–15 whereas the previously reported case of post-transplantation PJP was on ECMO support for 11 days.⁴ Compared with the average duration of PJP in transplant recipients and the described cases of ECMO for PJP, the duration of ECMO support in this case was very long, which even caused us to consider a lung transplantation.

As this case shows, recovery of pulmonary function is possible, even after longer periods of ECMO support.

Learning points.

• Pneumocystis jirovecii pneumonia (PJP) can cause severe respiratory failure in renal transplant recipients.

• Veno-venous extracorporeal membrane oxygenation (V-V ECMO) and, if needed, intermittent prone positioning can be a useful supportive therapy in patients with PJP.

• During V-V ECMO support, a protective ventilator strategy is very important. Ultraprotective ventilation strategies have been shown to be beneficial.

• Although symptoms of PJP in renal transplant patients typically have a short duration, there is no given time frame for V-V ECMO support; prolonged support may be necessary to allow for return of pulmonary function.