|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Division of Pulmonary Sciences and Critical Care Medicine (Dr. Dexter), University of Colorado at Denver and Health Sciences Center, National Jewish Research and Medical Center (Dr. Cosgrove), and Denver Health Medical Center (Dr. Douglas), Denver, CO.
Correspondence to: Ivor S. Douglas, MD, Denver Health Medical Center, 777 Bannock St, MC 4000 Denver CO, 80204; e-mail: idouglas{at}dhha.org
Key Words: extracorporeal support hypoxemic respiratory failure pulmonary alveolar proteinosis whole-lung lavage
A47-year-old woman was transferred from an outside hospital for evaluation of progressive hypoxia. She was in her usual state of health until 8 months prior to transfer, when a flu-like illness developed from which "she never recovered." Her initial predominant symptom was progressively worsening dyspnea on exertion; a dry cough developed later. She also reported episodes of palpitations associated with coughing, weight loss (subjective), and night sweats. She denied fever, hemoptysis, rash, and easy bruising/bleeding.
The patient had been repeatedly evaluated by an outside pulmonologist over the preceding 2 months, had undergone bronchoscopy twice with failure to diagnose the underlying condition, and was being treated empirically for an interstitial lung disease vs fungal infection at the time of transfer. Her transfer was arranged to the surgical service with a plan for video-assisted thoracic surgery biopsy at the time of her arrival for diagnosis.
Her medical history included allergic rhinitis as well as a single episode of community-acquired pneumonia 11 years previously. Her surgical history was noncontributory. She reported allergies to aspirin, ibuprofen, amoxicillin, and iodine (reported to cause respiratory arrest). Her medications on transfer were as follows: oxygen, nebulized ipratropium/albuterol qid; itraconazole, 200 mg bid, for possible fungal pneumonia; and prednisone, 60 mg/d; as well as cyclophosphamide, 75 mg/d, for presumptive interstitial lung disease with prophylactic double-strength trimethoprim-sulfa three times a week.
She reported a positive smoking history of 35 to 50 pack-years but quit early in 2005 due to respiratory illness. She denied any exposures to birds, air conditioners, hot tubs, or evaporative coolers. Her mother had died of tuberculosis, and there was no family history of rheumatologic or other pulmonary diseases.
On transfer, her vital signs were as follows: temperature, 35.3°C; pulse rate, 72 beats/min; BP, 90/56 mm Hg; respiratory rate, 18 breaths/min; and pulse oxygen saturation, 90% on a 15-L reservoir nasal cannula. She was a well-developed, well-nourished white woman in no apparent distress without significant conversational dyspnea. Pertinent positive examination findings included normal thoracic excursions with dry crackles bilaterally two thirds up the posterior chest, and additional frequent inspiratory squeaks audible on auscultation. She had decreased fremitus at the bases bilaterally without dullness to percussion. Her heart, abdominal, and extremity examinations were within normal limits. Her skin over the distal phalanges was thickened but not edematous, with loss of skin wrinkles in these same areas. She had no rashes or subcutaneous nodules.
What additional studies were performed?
A blood chemistry panel, liver function, and CBC findings were all within normal limits.
Radiography
Chest radiography showed diffuse bilateral alveolar infiltrates that were more hilar in predominance and showed diaphragmatic and perimediastinal sparing (Fig 1 ). High resolution chest CT demonstrated an extensive multilobar "crazy-paving" pattern with thickening of the interlobular septae, superimposed on ground-glass opacification in combination with patchy consolidation (Fig 2 ).
|
|
The patient continued to have increasing oxygen requirements over the next several hours that were not effectively treated by noninvasive positive pressure ventilation, resulting in endotracheal intubation. She underwent surgical biopsy, and on returning from the operating room required continued mechanical ventilation (MV) support due to profound hypoxemia. Bronchoscopic aspiration and lavage returned milky, thick fluid from the entire bronchial tree.
What working diagnoses were considered at this point?
A provisional diagnosis of idiopathic (acquired) pulmonary alveolar proteinosis was made based on the patients history (no relevant exposure history, no infections found on BAL or surgical biopsy, no recent immunosuppression prior to symptom onset) the radiologic crazy-paving appearance on chest CT scan, as well as the bronchoscopic and pathologic results.
Over the next 2 days, the patient remained severely and progressively hypoxemic with a PaO2 of 50 mm Hg on ventilator settings of fraction of inspired oxygen (FIO2) of 100% and increasing levels of positive end-expiratory pressure (PEEP) [15 to 22 cm H2O].
What was her clinical course and management?
Due to the patients progressive hypoxia and the recognition that her condition would not resolve without definitive therapy, she was scheduled for bilateral whole-lung lavage.
It was clear that single-lung ventilation would not provide sufficient oxygenation for the long procedure. As her condition was believed to be reversible over the short term with definitive intervention (whole-lung lavage [WLL]), and the clinical experience of one of the consulting faculty was significant in the use of extracorporeal membrane oxygenation (ECMO) for patients with pulmonary alveolar proteinosis (PAP) in similar situations, it was decided to consult the cardiothoracic surgery team to place the patient on ECMO therapy to augment her tissue oxygenation during the prolonged procedure. The risks, benefits, and severity of the situation were discussed with the patients husband, who wished to proceed.
After successful initiation of venovenous ECMO therapy using preprocedure heparinization and mild volume resuscitation to maintain sufficient flow through the circuit, the patient had improved oxygenation on flows of approximately 2 L/min. In conjunction with full ventilator support, her arterial oxygen saturation (SaO2) improved from < 80 to > 90%. She was then transitioned to single-lung ventilation on the left with decreased saturations from 80 to 85%. Her right lung was lavaged successfully in slowly increasing aliquots (400 mL to maximum of 1,000 mL), requiring a total of 12 L of warm normal saline solution with approximately 11 L of volume returned. There was dramatic clearing of the fluid from initial to final aliquot recovery. The patient was then changed to single-lung ventilation on the right with subsequent lavage of the left lung. Again, aliquots of warmed normal saline solution were infused and recovered in increasing volumes (400 mL to maximum of 750 mL) with significant clearance of proteinaceous fluid after 11 L were infused with approximately 10 L recovered. The procedure, in total, took approximately 3 h without significant complication. The SaO2 remained > 75% through the majority of the procedure, and she was hemodynamically stable throughout. At the end of the procedure, with bilateral lung ventilation and adjunctive ECMO, she had an SaO2 of 94% on an FIO2 of 100% and PEEP of 12 cm H2O. Due to her continued oxygen requirements, she was maintained on ECMO and was transported to the surgical ICU for further supportive care and monitoring.
Case Discussion
PAP: Pathogenesis, Evaluation, and Management
PAP is a rare disorder characterized by the accumulation of lipoproteinaceous material within the alveolar compartment.1 Three distinct forms of PAP exist: secondary PAP, congenital PAP, and acquired or idiopathic PAP. Environmental exposures, hematologic or neoplastic disorders, and immunodeficiency or immunologic disorders resulting in functional impairment or reduced numbers of alveolar macrophages are associated with the development of secondary PAP.234567891011121314151617 Pneumocystis jiroveci, cytomegalovirus, HIV, mycobacterial, and nocardial infections are associated with PAP, but it is unclear if the infections are causative or a result of the altered macrophage function in patients with idiopathic PAP.2345678 In congenital PAP, mutations in the gene encoding surfactant protein B, most commonly a frame-shift mutation at 375 termed 121ins2, occur. Premature termination of transcription occurs, leading to surfactant protein B deficiency.9 In addition to surfactant protein-B mutations, defects in granulocyte-monocyte colony stimulating factor (GM-CSF) receptor ß subunit gene have been reported in congenital PAP.10 Mutations in surfactant protein B and GM-CSF receptor ß subunit genes are not present in all patients with congenital PAP, suggesting that additional pathogenic mechanisms exist and further research is required.11
Transgenic mouse models have provided insight into the pathogenesis of acquired PAP. GM-CSFdeficient mice have an identical pulmonary phenotype to human patients with PAP.12 Adult human studies18 have not demonstrated mutations within the GM-CSF gene; rather, neutralizing autoantibodies to GM-CSF have been detected in virtually all patients with acquired PAP. These autoantibodies are a specific and sensitive marker of the disease. It therefore has been suggested that acquired PAP may be autoimmune in nature.1013
Acquired PAP is rare, with a prevalence of 0.37 per 100,000 persons. Of the three different forms, acquired PAP accounts for > 90% of cases.14 Mean age at the time of diagnosis is 39 years; 71% of patients are men.15 Patients generally present with progressive dyspnea, cough, and fatigue.16 Fever and weight loss may also complicate the course and may indicate a concomitant infection. Examination findings are nonspecific and often unremarkable. Clubbing is uncommon in PAP. Chest radiography typically demonstrates alveolar consolidation and ground-glass opacification, often in a perihilar or "bat wing" distribution. In more severe cases with diffuse opacification, there is often a notable diaphragmatic and/or perimediastinal sparing. The characteristic though nonspecific crazy-paving pattern on high-resolution CT reflects reticular opacities due to thickening of the interlobular septae, superimposed on ground-glass opacification, and is often seen in combination with patchy consolidation (Fig 2). The radiologic differential diagnosis of a crazy-paving pattern is summarized in Table 1 . Acute interstitial pneumonia, ARDS, pulmonary hemorrhage, and P jiroveci pneumonia are among the more frequent alternative explanations.171920
|
|
The postulated role for a GM-CSFdependent mechanism has prompted trials1213303132333435 assessing the efficacy of subcutaneous and inhaled GM-CSF to avoid WLL. Average response rates are 48% using subcutaneous GM-CSF (5 to 9 mg/kg/d).153233 In a recent retrospective study,34 inhaled GM-CSF (250 µg bid every other week for 12 weeks) was safe and effective in those patients with acquired PAP. In the same study,34 a response rate of 91% (11 of 12 patients) was observed, significantly higher than that reported for subcutaneous GM-CSF, with improvement in clinical, physiologic, gas exchange, and radiographic abnormalities. Following the discontinuation of inhaled GM-CSF, 41% (5 of 12 patients) had a symptomatic relapse, on average within 6 months, but responded to reinstitution of treatment. A durable response to therapy was noted in 4 of the initial 12 patients, up to 12 to 41 months of following the discontinuation of therapy.
Initial MV for Intractable Hypoxemic Respiratory Failure
Like many patients with intractable acute hypoxemic respiratory failure (AHRF) complicating diffuse lung disease (DLD), management of this patient with PAP presented a potentially insurmountable therapeutic challenge. Intractable severe hypoxemia is a less common direct cause of death than extrapulmonary multiple organ dysfunction in DLD including acute lung injury/ARDS. A primary differentiating factor in PAP is that AHRF will inexorably progress without definitive management by WLL. Thus, lung-protective MV was initiated with the intention of reducing the risk of superimposed ventilator-induced lung injury pending definitive management.
Although some PAP patients can be efficiently oxygenated assuming a permissive hypercapnic ventilatory approach by volume-cycled assist control MV, this was of only limited benefit in this patient. A recommendable lung-protective strategy is to maintain low tidal volumes of 5 to 6 mL/kg of predicted body weight, apply "optimal" PEEP to maintain plateau airway pressures < 30 cm H2O, and FIO2 targeted to pulse oxygen saturation of 88 to 92%. Chest CT may be clinically useful in determining the presence of PEEP-recruitable units and in avoiding overdistension.36 Decelerating wave form/high peak-inspiratory flow combinations with frequent monitoring for intrinsic (auto) PEEP was targeted. Pressure-cycled inverse ratio ventilation (inspiratory/expiratory ratio > 2:1) may potentially enhance recruitment but is associated with intrinsic PEEP accumulation. Diligent efforts were required to facilitate expiratory emptying using deep chemical sedation in an effort to avoid patient/ventilator dysynchrony and limit the further parenchymal complications of ventilator-induced lung injury.
As in this patient, routine modes of MV may prove unsuccessful, and numerous alternative modes and ancillary strategies have been proposed. Many of these interventions are poorly validated, having been applied in an anecdotal fashion and predominantly in desperately ill patients in the final stages of ARDS-associated AHRF with a low likelihood of recovery and not in patients with PAP. These patients are at substantial risk for complications of ventilation and critical illness including ventilator-associated pneumonia, prolonged chemical sedation, and extrapulmonary organ dysfunctions. Some of these modes include airway pressure release ventilation, high-frequency oscillatory ventilation, and supplemental high-pressure lung recruitment maneuvers and management in the prone position. Also reported, but probably of limited benefit, is the use of pharmacologic therapies to improve gas exchange. Surfactant dysfunction and deficiency is a central component in DLD-associated AHRF. Even in PAP characterized by abnormal surfactant clearance by alveolar macrophages, the presence of immature forms of surfactant protein B and insufficient palmitoylation of surfactant protein C contribute to dysfunction of the surfactant flooding the alveoli with progressive alveolar instability and progressive inflammation.37 Newer synthetic surfactant protein/lipid formulations may provide more affordable and feasible therapeutic options.38 Finally, despite clear evidence39 suggesting that oxidant stress plays a pivotal role in the evolution of AHRF, antioxidant therapies have proven disappointing.
Extracorporeal Support in PAP
Extracorporeal support for intractable AHRF is an infrequently applied technology. Since the first successful use of ECMO in 1972,40 there have been significant developments in the application of this technology as well as understanding of how best to avoid the inherent complications. ECMO has been used in the setting of cardiac and/or respiratory failure. In its most basic form, the apparatus removes carbon dioxide from venous blood, usually replacing it with oxygenated blood to be returned to the systemic circulation by either a venous or arterial route.
Early experience with ECMO prompted the National Institutes of Health to sponsor a prospective, randomized, multicenter trial41 of ECMO for ARDS in adult patients in 1975; this trial yielded a very low survival rate (10%) in both the treatment and control groups and effectively halted further research into this treatment modality for decades. In-depth analysis revealed that the study was methodologically flawed. There was an heterogenous experience with ECMO at the nine study sites; the enrollment criteria excluded the best-risk and worst-risk patients and included many patients who had influenza during a nationwide epidemic in 1976; there were significant bleeding complications > 2 L of blood loss per day; and most importantly a lung-protective ventilation strategy was not adopted, despite the primary intention of ECMO being to rest the injured lung. At postmortem lung examination, fixed fibroproliferative changes predominated, suggesting that the progression of the underlying disease may have contributed significantly to the failure to detect a clinical benefit.
Evidence from the intervening 30 years supporting the use of extracorporeal support modalities in adults continues to be mixed and of questionable validity. ECMO may be assistive in patients with severe, acute, potentially reversible respiratory failure that cannot be managed with optimal standard therapies.28424344
Despite these findings, several pediatric investigators45 continued to explore the use of ECMO for neonatal respiratory failure unresponsive to other management modalities, resulting in ECMO becoming a usual-care practice in this setting. Given the overall improvement in critical care management of patients and comparable survival for adult (53%) and pediatric ECMO patients (56%) with respiratory failure included in the Extracorporeal Life Support registry of 29,000 patients as of June 2004,45 it is reasonable to hypothesize that survival for adults might be improved if a similar study were to be performed today. The findings of the 180-patient, UK Conventional Ventilation or ECMO for Severe Adult Respiratory Failure Trial,46 which randomized adults with severe acute respiratory failure to conventional ventilation or ECMO, are anticipated in late 2007 and may provide clarity on the efficacy and safety of this modality in current critical care environments.
While acknowledging the absence of convincing evidence for safety or efficacy of ECMO and appreciating that alternative strategies including alternative MV modes and pulmonary vasodilators might have proven efficacious, we elected to initiate ECMO for our young, previously healthy patient with PAP. The availability of an ECMO-experienced supervising clinician was influential in this decision.
Since current guidelines suggest that prolonged MV (5 to 10 days) is a relative contraindication to ECMO due to a recognized increase in mortality,47 ECMO must be considered early in the course of intractable AHRF. Patients who are not able to undergo systematic anticoagulation should not be considered for this therapy. Investigators43 at the University of Michigan who have some of the most extensive published experience with this modality in adult patients have established relative contraindications to the use of ECMO for refractory hypoxemia (defined as PaO2/FIO2 ratio < 100 on FIO2 of 1.0, alveolar-arterial gradient > 600 mm Hg, or transpulmonary shunt fraction > 30% despite and after optimal treatment). These contraindications are age > 50 to 70 years and > 7 to 10 ventilator-days. Severe sepsis is no longer considered a contraindication for ECMO.48
A detailed discussion of methods for initiating and maintaining ECMO is beyond the scope of this review. Venovenous bypass is the preferred method in most centers, as it avoids brain hypoperfusion, maintains pulmonary blood flow, improves myocardial oxygenation, preserves the carotid artery, avoids systemic embolization, and avoids the possibility of arterial stenosis at insertion sites.49 Venovenous bypass relies on a stable intrinsic cardiac output and therefore is not traditionally a viable option in cardiac failure patients. Additionally, pumpless arteriovenous systems provide less efficient oxygenation than the more traditional venoarterial pumping devices50 but may provide a clinically viable future alternative.51
Clinical Follow-up
Despite severe respiratory distress, the patient had a salutary response to MV support with ECMO and WLL. Two days after the WLL procedure, the patient had significant improvement in oxygenation (PEEP of 7 cm H2O and FIO2 of 60%) and ECMO circuit decannulation was achieved without complication. After 3 additional days of MV, she was successfully extubated and discharged home on the fourteenth hospital day with 5 to 6 L oxygen by nasal cannula.
The patient was ultimately stable breathing room air after approximately 10 days at home. On reevaluation 3 months after discharge, she had experienced complete resolution of symptoms, and a repeat chest CT revealed only rare subpleural nodules (Fig 4 ). This salutary anecdotal experience with ECMO as a bridge therapy for intractable hypoxemic respiratory failure in this and similar patients prompts us to join an emerging call for new National Institutes of Health-funded prospective trials of ECMO in adults.
|
Footnotes
Abbreviations: AHRF = acute hypoxemic respiratory failure; DLD = diffuse lung disease; ECMO = extracorporeal membrane oxygenation; FIO2 = fraction of inspired oxygen; GM-CSF = granulocyte-monocyte colony stimulating factor; MV = mechanical ventilation; PAP = pulmonary alveolar proteinosis; PEEP = positive end-expiratory pressure; SaO2 = arterial oxygen saturation; WLL = whole-lung lavage
The authors declare that they have no conflicts of interest with regard to the content of this article.
Received for publication June 24, 2006. Accepted for publication October 10, 2006.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |