Pharmacotherapy for Autoimmune Pulmonary Alveolar Proteinosis

Abstract

Pulmonary alveolar proteinosis is suspected when a “crazy paving” pattern is observed on a chest CT scan. This diagnosis is confirmed by the presence of eosinophilic extracellular material that shows positive staining with Periodic Acid Schiff on bronchoalveolar lavage samples. The autoimmune form of pulmonary alveolar proteinosis is confirmed by detecting anti-granulocyte-macrophage colony-stimulating factor antibodies in the patient’s serum. The historical first-line treatment for autoimmune pulmonary alveolar proteinosis is whole lung lavage, which should only be performed in expert centers. It remains the preferred treatment for patients experiencing respiratory failure, especially at the time of diagnosis. Inhaled granulocyte-macrophage colony-stimulating factor supplementation with molgramostim or sargramostim is now considered a first-line treatment in the international guidelines for autoimmune pulmonary alveolar proteinosis, following the positive results of recent randomized placebo-controlled studies. Rituximab and plasmapheresis can be prescribed as third- and fourth-line treatments, respectively. Lung transplantation may be considered for eligible patients experiencing terminal respiratory failure. A deeper understanding of the pathogenesis of autoimmune pulmonary alveolar proteinosis has opened up new therapeutic avenues, such as the use of PPARγ agonists or statins.

Key Points

Management of autoimmune pulmonary alveolar proteinosis, from diagnosis to treatment, should be performed in expert centers.

Whole lung lavage and inhaled granulocyte-macrophage colony-stimulating factor supplementation are now positioned both as first-line treatment, possibly one after the other, especially in patients with respiratory failure.

In third- and fourth-line treatments are respectively positioned rituximab and plasmapheresis.

As surfactant components accumulated in pulmonary alveoli include esterified cholesterol, statin therapy may be efficient in patients with autoimmune pulmonary alveolar proteinosis.

Introduction

Pulmonary alveolar proteinosis (PAP) is characterized by the gradual accumulation of lipoproteinaceous surfactant material in the pulmonary alveoli, leading to impaired gas exchange [1, 2]. It is a rare disease, with an estimated prevalence of 7 cases per million people. The average age of those affected is around 50 years, with an equal sex distribution in the most recently published articles [3, 4].

Various forms of PAP exist, each associated with distinct pathophysiological mechanisms. Several classifications of PAP have been proposed [1, 2, 5]. For this article, we have selected a pragmatic classification approach:

- autoimmune PAP (aPAP), which is characterized by the presence of serum anti-granulocyte- macrophage colony-stimulating factor (GM-CSF) antibodies, and accounts for 90% of all PAPs, the focus of the present article;

- secondary PAP, associated with hematological disorders, malignant and non-malignant (mainly myelodysplastic syndromes), with chronic inhalation of toxic substances including indium, aluminum, titanium, silica, or cellulose fibers, and with chronic infections or immune deficiencies;

- genetic PAPs, which primarily affect children, and include MARS1-related PAP, mutations in GM-CSF macrophage receptor genes, and genetic mutations in surfactant protein SFTPB and SFTPC (encoding pulmonary surfactant-associated protein B [SP-B] and surfactant-associated protein B [SP-C]), ABCA3 (encoding ATP-binding cassette subfamily A member 3), and NKX2-1 (encoding thyroid transcription factor 1) [2]. In this article, we explore the pathophysiology of aPAP, outline diagnostic strategies, and describe both current and emerging therapeutic options.

Pathophysiology of Autoimmune Pulmonary Alveolar Proteinosis (aPAP)

Surfactant is a mixture of proteins and lipids, primarily composed of diphosphatidylcholine, which is synthesized and secreted by type II pneumocytes [6]. The four main surfactant proteins (SPs) are SP-A, SP-B, SP-C, and SP-D, which are encoded by the SFTPA1, SFTPA2, SFTPB, SFTPC, and SFTPD genes, respectively. Surfactant lipids are stored as intracytoplasmic lamellar bodies. Under physiological conditions, the components of surfactant are recycled by type II pneumocytes or catabolized by alveolar macrophages. These surfactant lipids and proteins play a crucial role in reducing alveolar surface tension, thereby preventing collapse during expiration. Additionally, they are involved in innate pulmonary defense mechanisms, particularly through SP-A and SP-D, which are members of the collectin family [6].

Granulocyte-macrophage colony-stimulating factor plays a major role in the pathophysiology of aPAP [7–10]. Granulocyte-macrophage colony-stimulating factor is a small glycoprotein involved in the regulation of alveolar macrophage differentiation. Autoimmune pulmonary alveolar proteinosis is induced by disruption of GM-CSF signaling, for example, when anti-GM-CSF autoantibodies are produced. Anti-GM-CSF antibodies can be detected in both the blood and broncho-alveolar lavage (BAL) fluid of patients with aPAP. These antibodies are pathogenic, they inhibit circulating GM-CSF, which in turn limits the activation of alveolar macrophages. As a result, these macrophages lose their ability to recycle surfactant leading to its accumulation in the pulmonary alveoli, and ultimately causing PAP.

Additionally, studies conducted on both humans and mice suggest that the GM-CSF–PU.1–PPARγ–ABCG1 pathway in alveolar macrophages plays a crucial role in maintaining surfactant homeostasis. Granulocyte-macrophage colony-stimulating factor has been shown to regulate cholesterol efflux from macrophages in a constitutive, dose-dependent, and reversible manner; however, it is not essential for the clearance of surfactant phospholipids [2].

Diagnosis

Diagnosis Strategy

Symptoms in patients with PAP are often non-specific and may include chronic cough, which can be accompanied by sputum production, as well as progressively worsening exertional dyspnea, which is the main reported symptom. The presence of fever or hemoptysis may indicate a complication, particularly a pulmonary infection. Chest pain could also be described by a patient, with or without infection.

Imaging plays a key role in PAP diagnosis. A chest X-ray can reveal interstitial lung disease. A chest CT scan is the primary tool for suspecting PAP, showing a characteristic ‘crazy paving’ pattern (Fig. 1) characterized by focal areas of ground-glass opacities with superimposed interlobular and intralobular reticular lines according to the Fleischner Society [11]. However, crazy paving is a non-specific finding that can be seen in a number of conditions (Table 1).

Fig. 1.

Autoimmune pulmonary alveolar proteinosis in a 46-year-old male smoker, with a 3-month history of cough and dyspnea without fever. Bilateral multilobar crazy paving areas (arrows) correspond to ground-glass superimposed with reticular images. Note the relatively well-defined boundaries between crazy paving and a healthy lung. Panels A, B, and C show axial slices at different levels of the lungs, panel D shows a coronal slice

Table 1.

Most frequent differential diagnoses of pulmonary alveolar proteinosis in the presence of a crazy paving pattern on a chest CT scan (adapted from Jouneau et al. [68])

Infections	Pneumocystosis pneumonia SARS-CoV2 infection (COVID)
Cancers	Invasive mucinous lung adenocarcinoma MALT lymphoma
Other ILDs	Alveolar hemorrhagea Exogenous lipid pneumonia

COVID coronavirus disease, ILDs interstitial lung diseases, MALT mucosa-assisted lymphoid tissue, SARS-CoV2 severe acute respiratory syndrome coronavirus 2

^aAssociated or not with connective tissue diseases

To confirm the diagnosis of PAP, a BAL should generally be conducted. The macroscopic appearance of the BAL fluid is typically described as milky or cloudy. Cytological analysis of the BAL fluid can confirm the diagnosis of PAP if it shows an accumulation of eosinophilic extracellular material, which will exhibit positive staining with Periodic Acid Schiff. Finally, the autoimmune nature of PAP is confirmed by the presence of anti-GM-CSF antibodies in the serum of patients. Lung histology, obtained through a surgical lung biopsy or cryobiopsy, is now rarely performed and even not recommended [5]. This diagnosis algorithm is summarized in Fig. 2.

Fig. 2.

Algorithm for diagnosis of patients with autoimmune pulmonary alveolar proteinosis (adapted from European Respiratory Society guidelines for the diagnosis and management of pulmonary alveolar proteinosis, McCarthy et al. [5]). GM-CSF granulocyte-macrophage colony-stimulating factor

Prognosis and Complications

No death was reported in the 5-year period of this Japanese national registry-based study including 223 patients with aPAP [3]. The survival rate of patients with all types of PAP is 78% at 2 years, 75% at 5 years, and 68% at 10 years in a cohort of 343 patients, including mainly “idiopathic PAP,” which is the old name for aPAP [12]. In this series, 69 patients died and 80% of deaths occurred within the first year of diagnosis. The majority of deaths were attributable to PAP with 72% due to progression to chronic respiratory failure and 18% due to infections [12]. A recent publication analyzed 3278 patients with PAP from 295 studies, including 2903 (88.6%) with idiopathic PAP [13]. The mean follow-up was 1.8 ± 1.6 years and the mortality rate was 6.8% in idiopathic PAP. The two major causes of death in non-secondary PAP (n = 231) were respiratory failure (108/231, 46.8%), and respiratory tract infections (36/231, 15.6%).

The clinical course of lung disease in aPAP follows one of three patterns [2]: progressive deterioration, stability, or spontaneous resolution. In a large Japanese study involving 223 patients with aPAP, the duration of the disease significantly influenced its progression [3]. Among individuals recently diagnosed with aPAP (less than 1 year since onset), two thirds reported that their symptoms remained unchanged from the time of diagnosis to the time of registration. For patients with an intermediate duration of aPAP (1–10 years), 42.5% experienced an improvement, 29.8% reported worsening symptoms, and 27.7% remained stable. In patients with a prolonged duration of aPAP (more than 10 years), nearly two thirds experienced a decline in their condition [3]. Spontaneous improvement may occur in 8–25% of cases [3, 12, 14]. Complete resolution after cessation of exposure (dust or tobacco) has been described in 2.7–10% of patients [3, 13, 14].

Several studies show that aPAP may evolve towards pulmonary fibrosis [15–17]. In a recent publication of the French aPAP cohort, signs of fibrosis, mainly traction bronchiectasis and bronchiolectasis, were found in 26% of patients after a median follow-up of 3.6 years [18]. The risk of fibrosis was higher in cases of dust exposure, and was associated with reduced survival.

Infectious complications should be investigated at PAP diagnosis, and during follow-up in the case of fever and an unexplained clinical or radiological deterioration such as the occurrence of pulmonary consolidation or nodules [3, 12]. Half of patients do not have fever during infectious episodes [19]. The main infections associated with PAP may be secondary to common microorganisms such as streptococci, Haemophilus spp. or enterobacteria, but they may also be secondary to opportunistic microorganisms such as mycobacteria, Nocardia spp., Actinomyces spp., Aspergillus spp., or cryptococci [12, 13, 19, 20]. In the French series of patients with aPAP diagnosed between 2008 and 2018, Nocardia spp. was the main micro-organism found, in 10 out of 104 patients, half of them at the diagnosis of aPAP, and half of them with cerebral abscesses [20].

Therapeutic Management

First and foremost, patients with aPAP should be managed after discussion with, or referral to, an expert PAP center. Treatment of patients with aPAP is based on:

• lifestyle modifications, including smoking cessation and avoidance of dust inhalation;

• general measures, including vaccinations and symptomatic treatments designed to improve respiratory symptoms;

• and more specific aPAP therapies.

A treatment algorithm is proposed in this article, adapted from European Respiratory Society (ERS) international guidelines (Fig. 3) [5].

Fig. 3.

Algorithm for management of patients with autoimmune pulmonary alveolar proteinosis (adapted from European Respiratory Society guidelines for the diagnosis and management of pulmonary alveolar proteinosis, McCarthy et al. [5]). GM-CSF granulocyte-macrophage colony-stimulating factor

Lifestyle Modifications

As in the general population, and for any chronic respiratory disease, to avoid dust exposure is important. As mentioned earlier, complete resolution following the cessation of exposure to dust or tobacco has been reported in 2.7–10% of patients [3, 13, 14]. Thus, smoking cessation is essential; tobacco smoking is a risk factor for the recurrence of disease after whole lung lavage (WLL) [21].

Exposure to dusts or other airborne pollution can lead to (secondary) PAP and potentially to the occurrence or aggravation, of particularly pulmonary fibrosis, in patients with aPAP [1, 2, 18]. It is recommended that patients with aPAP limit their exposure to airborne pollution, especially dusts. This has to be done in collaboration with occupational medicine, and the adaptation of the work station or even professional reclassification.

General Management

Vaccinations against influenza, pneumococcus, and severe acute respiratory syndrome coronavirus 2 are recommended. Vaccination against respiratory syncytial virus, may be discussed before 1 year of age and after 65 years of age. Symptomatic treatments are those common to all chronic respiratory diseases: respiratory rehabilitation, oxygen supplementation, and palliative care.

Specific Treatments of aPAP

In the absence of spontaneous improvement, or in the case of severe disease at aPAP diagnosis, several treatments are available: WLL, which is the “historical” first-line and reference treatment, but recent studies on inhaled GM-CSF positioned this therapy also as first-line treatment (Fig. 3) [5]. Rituximab and plasmapheresis come in third and fourth positions, respectively. Lung transplantation is rare but possible. Other emerging treatments include statins and PPARγ agonists.

Whole Lung Lavage (WLL)

Historically, this has been the reference treatment. In the ERS guidelines, like GM-CSF supplementation, it is proposed as a first-line treatment, particularly in patients with severe diseases such as oxygen supplementation at rest or respiratory failure [1, 2, 5]. It should be performed in expert centers with pulmonologists and anesthesiologists/intensive care physicians familiar with this procedure.

The patient is placed supine, intubated with a selective double-lumen endotracheal tube under general anesthesia, and given a neuromuscular blockade. Correct positioning of the endotracheal tube is checked by flexible bronchoscopy. One channel of the probe is used to ventilate the patient (protective ventilation, FiO₂ = 100%, continuous capnography [EtCO2]), while the other channel is used for lung lavage. Saline at 37 °C is instilled until the lung is full, then the liquid is evacuated by gravity. Systematic microbiological sampling is recommended: bacteriology (standard, Nocardia, Actinomyces, mycobacteria), myco-parasitology (Aspergillus, Cryptococcus), and virology (severe acute respiratory syndrome coronavirus 2) on the first liter of effluent [20]. The procedure (instillation-removal) is repeated until the effluent liquid becomes clear (Fig. 4). On average, 15–20 L of saline is necessary to wash each lung, but it can be as much as 50 L. Whole lung lavage implies a bilateral procedure [5]. In patients with severe respiratory failure, WLL can be performed under venovenous extracorporeal membrane oxygenation; however, this is typically not necessary.

Fig. 4.

Effluent fluid from a whole lung lavage procedure in patients with autoimmune pulmonary alveolar proteinosis with heavy smoking (tobacco and cannabis). Note the deposition of lipoproteinaceous material in the bottom of the jars (yellow arrows), which decreases from right (beginning of whole lung lavage) to left (end of whole lung lavage), accompanied by a decrease in effluent turbidity

Numerous adaptations have been proposed to enhance the efficacy of WLL (i.e., increase the turbidity of the effluent): the possibility of adding concomitant thoracic percussion (manual or mechanical); the possibility of performing changes of position (alternating procubitus-decubitus) during WLL; or last, the possibility of a specific ventilatory mode (repeating periods of “balloon” ventilation during WLL). Successful segmental lavage under local anesthesia and flexible bronchoscopy has also been reported in patients with less severe forms.

Complications of WLL are rare when performed in an expert center. They include desaturation, prolonged mechanical ventilation, contralateral leakage, pneumothorax, subcutaneous emphysema, headache, convulsion, and hyperthermia, which may be associated with infection [22–24]. Mortality is extremely rare.

Indications for WLL are established by expert centers, and ERS guidelines “recommend performing bilateral whole lung lavage in patients with aPAP with evidence of gas exchange impairment and either symptoms, or functional impairment (strong recommendation, very low certainty of evidence)” (Fig. 3) [5]. Whole lung lavage should not be performed in cases of uncontrolled infection, owing to the risk of infectious dissemination or even septic shock [2].

Following WLL, there is an improvement in symptoms, in 68–90% of patients, and in exercise tolerance, imaging, partial pressure of oxygen, and pulmonary function tests (forced expiratory volume in 1 second, forced vital capacity [FVC], diffusing capacity of the lungs for carbon monoxide [DLCO]) [12, 23–25]. Single (bilateral) WLL is sufficient in half of the patients, suggesting that the presence of anti-GM-CSF antibodies is not sufficient to maintain disease [26]. In the other half of patients, it will be necessary to repeat WLLs, with an average of one more WLL (extremes: 1–22) [12]. Whole lung lavage is also associated with a reduced incidence of opportunistic infections [25].

Active smokers needed a higher number of WLLs. Indeed, it has been reported the need for an average of five WLLs in active smokers to achieve remission, compared with 2.4 in non-smokers [21]. The 5-year survival rate after these WLLs was 94 ± 2% in Seymour and Presneill’s 2002 review of the literature [12].

Regarding the cost of this treatment, it is influenced by several factors, including the cost of hospitalization, the type of unit (intensive care unit, operating room, or pulmonology ward), and the duration of the hospital stay. The average daily cost of hospitalization in the USA is approximately $3000, which means that the total cost for a bilateral WLL typically ranges from $15,000 to $20,000.

Regarding accessibility, not all expert centers for rare pulmonary diseases are able to perform WLL, which can complicate the organization of this treatment. Additionally, WLL is unavailable in some countries.

Regarding the risks associated with WLL, the side effects have been outlined above. The benefits of the procedure can be observed relatively quickly, which is why we recommend WLL for patients with more severe cases of aPAP.

Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF) Supplementation

Granulocyte-macrophage colony-stimulating factor supplementation saturates anti-GM-CSF antibodies and thus restores alveolar macrophage differentiation. Two routes of administration have been described: subcutaneous injection and inhalation. Two molecules exist: sargramostim, developed for the injectable route (but also used by the nebulized route), and molgramostim, developed specifically for inhalation. Studies published since 2019 suggest that this therapy should be discussed as a first-line treatment alternative to WLL.

Subcutaneous GM-CSF Several studies have described the benefit of subcutaneous injections of GM-CSF in aPAP. Three prospective studies used daily injections of 5–9 µg/kg/day of GM-CSF for 12 weeks, with doses gradually increasing to 20 µg/kg/day according to clinical responses [27–29]. Treatment was effective in almost half the patients: 43% in the longest study (1 year) and up to 75% in the shortest study (12 weeks) [27–29]. A maintenance dose for 1 year has been used successfully in a few patients (20 µg/kg, three times a week) [29]. Adverse effects of GM-CSF subcutaneous injections are minor: edema, erythema, and pain at the injection site, fever, chills, nausea, vomiting, malaise, headache, asthenia, arthralgia, and dyspnea [27–29]. Very rarely, neutropenia or eosinophilia may occur, reversible on discontinuation of treatment (recommended if neutrophils are below 1.5 G/L) [28, 30]. This modality is currently almost abandoned.

Inhaled GM-CSF Inhaled GM-CSF enables the drug to be delivered directly to the lungs, with less systemic toxicity. Doses can also be taken at longer intervals and, given the high cost of GM-CSF, the inhaled route is less expensive compared with subcutaneous injections. Moreover, the inhaled route is more effective than the subcutaneous injection route [30].

The first studies date back to the 2000s [31], but between 2019 and 2024, five randomized placebo-controlled studies have demonstrated the efficacy of inhaled GM-CSF in patients with aPAP (Table 2) [32–37]. The most recent meta-analysis including three randomized studies using inhaled GM-CSF showed a significant improvement in quality of life assessed by the St George Respiratory Questionnaire (SGRQ, mean difference − 8.09, 95% confidence interval [CI] − 11.88, − 4.3, p < 0.0001), distance to the Six-Minute Walking Test (mean difference 21.72 m, 95% CI − 2.76, 46.19, p = 0.08), DLCO (% of predicted : mean difference 5.09%, 95% CI 2.05, 8.13, p = 0.001,) and alveolo-arterial difference in oxygen (A-aDO2, mean difference − 4.36 mmHg, 95% CI − 7.19, − 1.52, p = 0.003) [38].

Table 2.

Summary of randomized clinical trials assessing inhaled GM-CSF versus placebo in patients with autoimmune pulmonary alveolar proteinosis

Author, journal, year, trial name	GM-CSF molecule	Dose and duration	After WLL	Number of patients	Efficacy
Tazawa et al., N Engl J Med, 2019, PAGE trial [34]	Sargramostim	125 µg ×2/jday for 7 days, every other week for 24 weeks	No	63 33 GM-CSF 30 placebo	Positive primary endpoint: significant decrease in AaDO2: 4.50 ± 9.03 mm Hg vs 0.17 ± 10.50 mm Hg, p = 0.02 Significant decrease in lung density on a chest CT scan No significant improvement in QoL (CAT questionnaire), FVC, DLCO, and 6MWT
Tian et al., Orphanet J Rare Dis, 2020 [35]	Molgramostim	150 µg ×2/day for 7 days, every other week for 3 months followed by 150 µg ×1/jday for 7 days, every other week for 3 months, for a total of 6 months (with an additional follow-up of 24 months)	No	36 19 GM-CSF 17 placebo	Negative primary endpoint: no significant decrease in AaDO2: 7.31 ± 8.81 mm Hg vs 1.8 ± 11.21 mm Hg, p = 0.146 (6 months) Significant improvement of QoL (SGRQ) at 6 and 24 months, TLC, and DLCO only at 24 months No efficacy on a chest CT scan or number of rescue WLLs
Trapnell et al., N Engl J Med, 2020, IMPALA trial [37]	Molgramostim	300 µg ×1/day either: daily (continuous arm), or for 7 days, every other week (intermittent arm) for 24 weeks	No	138 46 continuous GM-CSF 45 intermittent GM-CSF 47 placebo	Positive primary endpoint in continuous inhaled GM-CSF arm: significant decrease in AaDO2: − 12.8 mmHg vs − 6.6 mmHg; mean difference − 6.2 mm Hg, p = 0.03 Significant improvement of all other endpoints in the continuous arm: QoL (SGRQ), DLCO, a decrease in number of rescue WLLs, and a decrease in ground glass opacity score on a chest CT scan
Campo et al., Eur Respir J, 2024 [32]	Sargramostim	250 µg ×1/day for 7 days, every other week for 12 weeks followed by 4 weeks of washout, and then 250 µg ×1/jday on day 1 and day 3 every 14 days for 6 months For a total of 10 months	Yes (initiated within 7 days after WLL)	18 9 GM-CSF 9 placebo	Positive primary endpoint: increase of median time to rescue WLLs in GM-CSF arm compared with placebo: 18 (IQR 6, 27) months vs 30 (30, 30) months, p < 0.0078 Significant decrease in AaDO2, PaO2, and DLCO Non-significant decrease in lung density on a chest CT scan and non-significant improvement of QoL (SF-36)
Trapnell, N Engl J Med, 2025, IMPALA-2 trial [36]	Molgramostim	300 µg ×1/day for 48 weeks	No	164 81 GM-CSF 83 placebo	Positive primary endpoint: significant increase in DLCO at week 24: mean difference +6.0 mm Hg, p = 0.0007 Significant improvement of numerous other endpoints: DLCO at week 48 (change from baseline 11.60% vs 4.70%; estimated treatment difference, 6.90%; p = 0.0008), QoL (SGRQ), AaDO2, DSS, decrease in ground glass opacity score on a chest CT scan, improvement of exercise capacity on a treadmill

6MWT Six-Minute Walk Test, AaDO2 alveolo-arterial difference in oxygen, CAT questionnaire COPD assessment test questionnaire, DLCO diffusion of carbon monoxide, DSS disease severity score, FVC forced vital capacity, GM-CSF granulocyte-macrophage colony-stimulating factor, IQR interquartile range, PaO₂ partial pressure of oxygen, QoL quality of life, SGRQ St George Respiratory Questionnaire, TLC total lung capacity, WLL whole lung lavage

The PAGE trial randomized patients with aPAP to receive either inhaled recombinant human GM-CSF sargramostim at a dose of 125 µg twice daily for 7 days every other week over a 24-week period, or a placebo [34]. The primary endpoint was met, demonstrating a significant change in the mean (± standard deviation) A-aDO2, which improved more in the GM-CSF group (33 patients) compared with the placebo group (30 patients): mean change from baseline was − 4.50 ± 9.03 mmHg versus 0.17 ± 10.50 mmHg; p = 0.02. Additionally, lung density on chest CT scans was significantly reduced in the sargramostim group compared with the placebo group. However, there were no significant improvements observed in other endpoints, including quality of life (as measured by the COPD assessment test questionnaire), FVC, DLCO, or the Six-Minute Walking Test (Table 2).

The IMPALA(-1) trial randomly assigned patients with aPAP to receive the recombinant GM-CSF molgramostim (300 µg once daily by inhalation) either continuously or intermittently (every other week), or a matching placebo for a duration of 24 weeks [37]. The primary endpoint was met, showing a greater improvement in the A-aDO2 among patients receiving continuous molgramostim (n = 46) compared with those receiving placebo (n = 47): − 12.8 mm Hg versus − 6.6 mm Hg; estimated treatment difference of − 6.2 mm Hg; p = 0.03. Furthermore, numerous secondary endpoints were significantly improved in the continuous inhaled molgramostim group compared with the placebo group, including quality of life (as measured by the SGRQ), DLCO, and a decrease in the ground-glass opacity score on chest CT scans (Table 2).

The positive IMPALA-2 trial recently published [36], with the largest number of patients with aPAP included (n = 164), brings even more evidence on the efficacy of inhaled GM-CSF in patients with DLCO < 70% of predicted values [36]. The primary endpoint was positive with a significant increase in DLCO at week 24: mean difference +6.0 % of predicted, p = 0.0007. There was also a significant improvement of all other endpoints: DLCO at week 48 (change from baseline 11.60% vs 4.70%; estimated treatment difference, 6.90%; p = 0.0008), quality of life (SGRQ), AaDO2, disease severity score, a decrease in the number of rescue WLLs, and a decrease in the ground glass opacity score on a chest CT scan. Finally, the mean change from baseline in exercise capacity using an exercise treadmill test increased more rapidly and was greater at week 48 for molgramostim than placebo [36].

In these studies, no particular adverse effects were observed. Despite the use of GM-CSF, blood leukocyte counts remained unchanged. No predictive factors for the efficacy of inhaled GM-CSF could be identified [34].

Therefore, the recent ERS guidelines “recommend inhaled GM-CSF as first-line treatment for symptomatic patients with confirmed aPAP (Strong recommendation, very low certainty of evidence)” (Fig. 3) [5]. Before publication of these international guidelines but after the end of a systematic literature review (updated on 9 August, 2022.), Campo et al. showed that administration of inhaled GM-CSF immediately after a WLL greatly reduced (by a factor of 7) the risk of relapse and the need for a new WLL in a 3-year follow-up of patients with aPAP [32]. Therefore, in the 7 days following a WLL, the initiation of inhaled GM-CSF may be discussed.

Regarding side effects, the overall incidence of adverse events and serious adverse events in these trials did not differ significantly between inhaled GM-CSF and placebo. However, certain adverse events were reported more frequently in the molgramostim group compared with the placebo group. For instance, in the IMPALA trial, chest pain occurred in 21.7% of patients receiving molgramostim versus 2.2% in the placebo group [37], although this was not observed in the IMPALA-2 trial. In the IMPALA-2 trial, diarrhea was reported in 11% of the molgramostim group compared with 2% in the placebo group, and coronavirus disease 2019 infections occurred in 22% of the molgramostim group versus 10% in the placebo group [36]. Notably, serious adverse events in the IMPALA-2 trial were numerically more frequent in the placebo group compared with the molgramostim group, with rates of 17% versus 24%, respectively.

Regarding the cost of this treatment, there is currently no established price for molgramostim, as this molecule has not yet received approval from the US Food and Drug Administration or European Medicines Agency. Regarding sargramostim, the cost varies depending on the website or pharmaceutical manufacturer, typically with a range from $300 to $400 per vial of 250 µg. For instance, based on the latest published protocol involving sargramostim (Table 2, Campo et al. [32]), a 10-month course of treatment would cost between $20,000 and $26,000.

In terms of accessibility, molgramostim is not yet available. However, sargramostim is widely accessible and may be provided under compassionate or temporary use programs.

Regarding the risks associated with inhaled GM-CSF, the side effects appear to be infrequent and manageable. Some patients may find the daily use of the inhaled treatment to be cumbersome. However, the benefits of inhaled GM-CSF are well established, supported by several positive randomized controlled trials. Additionally, compared with the historical gold standard of WLL, inhaled GM-CSF offers the advantage of avoiding hospitalization and general anesthesia.

Rituximab

A systematic review identified a single interventional study with a single arm involving ten patients [39], a retrospective series of 13 patients [40], and nine case reports, including only adults with aPAP. Most of these patients had previously received WLL or GM-CSF supplementation.

Data from the Kavuru et al. and Soyez et al. studies were pooled for comparison [39, 40]. Participants received two 1000-mg doses of rituximab, administered 15 days apart. Results suggested that rituximab could reduce A-aDO2, by 15.59 mmHg [95% CI 12.15,19.03], and improve partial pressure of oxygen by 19.6 mmHg [8.39, 30.81]. Results suggest little or no impact on DLCO + 16.8% [− 4.86%, 38.46%] of predicted value, FVC + 2.7% [− 22.47%, 27.87%], or the Six-Minute Walk Test +19 [− 93.47, 131.47] m). Kavuru et al. reported that 4/7 of patients observed for a mean of 32 (±6) months did not require WLL [39]. The remaining three patients required one WLL each during the follow-up. Soyez et al. reported that 4/11 patients showed a significant improvement at 12 months, defined as a decrease in A-aDO2 of at least 10 mmHg [40].

Kavuru et al. also reported a significant improvement in radiological scores (p = 0.027), but this was not observed by Soyez et al. No deaths or serious adverse events were observed in these studies. These results must be interpreted with caution because of the sample size and the absence of a control group.

Seven of the nine reported cases documented a clinically significant improvement between 3 and 12 months post-infusion: weaning from oxygen supplementation, improvement in oxygen saturation, improvement in exercise capacity, reduction in WLL frequency, or improvement in DLCO or FVC. Two of the nine cases (22.2%) reported no benefit from rituximab, although this proportion should be interpreted with caution because of publication bias. Moreover, spontaneous remission was observed in 8–25% of patients with aPAP, thus a therapeutic effect cannot be established with certainty on the basis of these data.

Although the safety of rituximab has not been specifically evaluated in patients with aPAP, abundant data are available in other disease areas. More specifically, the safety profile of rituximab at a similar dose (two 1000-mg doses) in adults has been assessed in greater detail in a Cochrane review evaluating rituximab in rheumatoid arthritis [41].

The pathophysiology of aPAP and the available data may therefore justify the prescription of rituximab as a third-line treatment, after the failure of WLL and inhaled GM-CSF and advice from a rare lung disease expert center (Fig. 3). The ERS guidelines “suggest the use of rituximab for patients with confirmed autoimmune PAP who remain symptomatic, requiring supplemental oxygen, despite whole lung lavage therapy or exogenous GM-CSF treatment (conditional recommendation, very low certainty of evidence)” [5].

Regarding side effects, most of the literature originates from rheumatology, particularly in the context of rheumatoid arthritis. The most significant side effects reported include immunosuppression due to B-lymphocyte depletion and an increased risk of infections. Publications evaluating rituximab in the treatment of aPAP have documented common adverse events, such as fatigue (observed in five out of ten patients), nasal congestion (also in five out of ten patients), and headache (reported by two out of ten patients) [39]. Regarding infections, two out of ten patients reported upper respiratory tract infections, and no other infections, particularly opportunistic infections, were documented. In the retrospective study conducted by Soyez et al., no serious adverse events related to rituximab were identified. Two patients developed viral upper respiratory tract infections, one patient was diagnosed with Mycobacterium avium airway colonization, and no opportunistic infections were observed. [40]. Since these publications, the coronavirus disease 2019 pandemic has highlighted the increased risk of coronavirus disease 2019 in unvaccinated patients receiving treatment with rituximab.

Regarding the cost of this treatment, rituximab is available in several biosimilar formulations, with prices that range from approximately $500 to $1000 per 1000-mg vial. Regarding accessibility, biosimilars of rituximab are widely available and are prescribed for a variety of conditions, including those in internal medicine, rheumatology, and hematology. If the treatment is not administered in a respiratory diseases department, it can be organized in other departments that are accustomed to performing these infusions.

Regarding the risk-benefit ratio, the immunosuppression induced by rituximab must be considered. Compared with first-line treatments, rituximab may be more cost effective and may be organized and administered more quickly, requiring only a few hours for infusions on days 1 and 15. This approach avoids the need for several days of hospitalization and general anesthesia, as is required for WLL, and eliminates the need for repeated inhalations of GM-CSF, as seen with sargramostim or molgramostim. However, the efficacy and level of evidence for rituximab are not as well established as those for WLL or especially inhaled GM-CSF.

Plasmapheresis

The data are based only on nine case reports [42–50]. All reported cases had severe disease with eight of nine receiving oxygen therapy up to 8 L/min and one patient mechanically ventilated. All had undergone multiple WLLs. Four of the nine patients had received GM-CSF and one had received rituximab previously. No significant clinical benefit was seen in three reported cases [45, 47, 50]. A modest response was reported in two cases. More specifically, Yu et al. [49] reported an improvement in symptoms and a CT scan, but with a relapse at 5 months. Luisetti et al. [48] reported a reduced incidence of WLL after plasmapheresis. Finally, four cases reported a significant improvement in symptoms [42–44, 46], oxygenation, CT scan, or pulmonary function tests [42, 46]. A significant decrease in anti-GM-CSF antibody titers was reported in five of nine patients [42–44, 48]. The pathophysiology of aPAP and the available data may therefore justify the prescription of plasma exchange after previous lines of treatment and advice from an expert center. The ERS guidelines “suggest the use of plasmapheresis for patients with confirmed aPAP who remain symptomatic, requiring high flow of supplemental oxygen (≥4L /min) or two or more WLL over a period of a year, despite receiving exogenous GM-CSF and rituximab, or having previously failed these treatments (conditional recommendation, very low certainty of evidence)” (Fig. 3) [5].

Regarding side effects, they are common but rarely severe. Episodes of low blood pressure occur in 20% of patients, although 95% of these cases are asymptomatic. Hypocalcemia is observed in 12% of patients, with 75% of these cases being asymptomatic. Additionally, no anaphylactic reactions have been reported [51].

Regarding the cost of this plasmapheresis, it varies depending on the country. The use of albumin alone requires approximately 4 L, which costs around $400. Additionally, five to ten sessions of plasmapheresis are needed, with each session costing about $250, resulting in a total cost of $1250–$2500 for the plasmapheresis kits alone. When factoring in hospitalization costs, the overall expense for the entire protocol ranges from approximately $3000 to $6000.

In terms of accessibility, plasmapheresis necessitates a skilled team, typically located in the intensive care unit of a tertiary center, owing to the specialized expertise of respiratory centers in diagnosing and managing aPAP.

Regarding the benefit-risk ratio, there is a lack of high-level evidence supporting the efficacy of this treatment for aPAP. Plasmapheresis may not only reduce anti-GM-CSF antibodies but may also exert additional immunomodulatory effects on B and T lymphocytes [52]. When comparing plasmapheresis with other treatments, WLL provides a quicker improvement for patients, while inhaled GM-CSF has a higher level of evidence supporting its efficacy. Additionally, rituximab is easier to administer and has more robust evidence of effectiveness, with 23 patients documented in two case series compared with only nine in case reports for plasmapheresis. These factors position plasmapheresis within the last therapeutic options for patients with aPAP. However, based on the availability and expertise of each center, plasmapheresis may occasionally be utilized earlier than currently recommended.

How to Select Between Treatments for aPAP?

First, as previously mentioned, smoking cessation and avoidance of personal or occupational exposure to airborne pollutants should be considered essential. Subsequently, we believe that the therapeutic algorithm (Fig. 3) should be adapted based on the patient’s condition, available resources, and local healthcare organization.

In patients with asymptomatic aPAP and limited functional or radiological disease, surveillance alone is recommended. In cases of severe aPAP with respiratory failure, WLL should be performed, potentially in conjunction with extracorporeal membrane oxygenation in the most critical situations. The initiation of inhaled GM-CSF within the subsequent 7 days should be systematically considered, taking into account the findings from the study by Campo et al. [32]. It is important to note that WLL should be conducted by an experienced team, and patients may need to be transferred to other centers, and in some cases, even to other countries.

In patients without respiratory failure but with significant disease—specifically, those who are symptomatic and have only mild-to-moderate gas exchange impairment—inhaled GM-CSF may be considered as a first-line treatment. The evidence supporting this therapeutic option is robust, as it is backed by several, consistent, randomized controlled clinical trials [32, 34, 36, 37]. However, these medications are expensive, and not all centers or patients may be able to afford them. Therefore, rituximab or plasmapheresis may be considered earlier in the therapeutic algorithm than indicated in Fig. 3, with a careful assessment of the benefit-risk ratio.

Lung Transplantation

Lung Transplantation The indication of lung transplantation may be considered for aPAP, as for other chronic respiratory diseases, for those with end-stage irreversible pulmonary fibrosis in addition to terminal respiratory failure despite optimal medical treatment, and in the absence of absolute contraindications to lung transplantation [53–55]. However, it remains extremely rare in the context of aPAP. Very few data are available concerning patients transplanted specifically for PAP, with only clinical cases [56, 57]. To date, a single case of recurrence of aPAP has been reported in a patient who underwent lung transplantation for aPAP with fibrosing evolution [58]. This patient was successfully treated with inhaled sargramostim. Notably, there have been reported cases of PAP developing after lung transplantation for conditions other than aPAP, particularly secondary PAP associated with the use of immunosuppressive medications [59, 60].

Statins In patients with aPAP, alveolar macrophages show a marked increase in cholesterol but only a slight increase in phospholipids, and pulmonary surfactant shows an increase in the cholesterol/phospholipid ratio. Statin treatment reduces cholesterol levels in alveolar macrophages in vitro. In GM-CSF receptor knockout mice (CSF2RB-/-), statin treatment reduces cholesterol accumulation in alveolar macrophages and improves PAP, and ex vivo statin treatment increases cholesterol efflux from macrophages [61].

Four clinical cases report radiological or functional improvement with statins [61–63]. A Taiwanese retrospective study reported no effect on survival in the 15 patients treated with statins in a cohort of 276 patients with PAP [64].

A prospective Chinese observational study included 47 patients with PAP without hypercholesterolemia who received an oral statin with a 12-month follow-up [65]. Forty patients completed the study, there was no significant difference in A-aDO2 but the authors report 26 (65%) patients including four patients with a complete response and 22 patients with a partial response on thoracic imaging, A-aDO2, and DLCO, without clear definition. Factors associated with a response were higher levels of anti-GM-CSF antibodies and baseline total cholesterol/high-density lipoprotein cholesterol. No serious adverse events were observed during the study.

Despite the methodological limitations of the latest study and the paucity of available data, the cost and well-known safety profile of statins may justify their prescription in addition to or following the treatments discussed above, after advice of an expert center. However, further prospective clinical studies are needed to determine the efficacy of statins in patients with aPAP.

PPARγ Agonists The pathophysiology of aPAP involves the GM-CSF receptor and the PPARγ and PU.1 pathways, which PPARγ agonists may selectively activate. In vitro/in vivo studies and one case report have reported improvements with pioglitazone [50, 66, 67]. The value of pioglitazone in the treatment of aPAP is currently the subject of a ongoing prospective study (NCT03231033).

Conclusions

Pharmacotherapy of aPAP has evolved considerably, and the ERS guidelines published in November 2024 propose for the first time a clear hierarchical treatment algorithm (Fig. 3) [5]. In all patients, smoking cessation and avoidance of airborne contaminants are crucial. In patients with mild disease, regular monitoring may be sufficient, as up to 25% of patients with aPAP may experience a spontaneous improvement.

Whole lung lavage, the historical first-line treatment, should now be discussed along with inhaled GM-CSF supplementation, given the numerous high-quality studies published in recent years. In the case of refractory aPAP, ERS guidelines clarify the therapeutic sequence that places rituximab as a third-line treatment and plasmapheresis as a fourth-line option. However, based on the patient’s preferences, local resources, costs, and accessibility to treatment, the sequence of care may be adjusted to ensure optimal management for patients with aPAP. New therapies are currently under evaluation, and it is likely that the diagnostic and therapeutic management of aPAP will continue to evolve in the future.

Funding

Open access funding provided by Centre Hospitalier Universitaire de Rennes. No funding was received for the preparation of this article.

Declarations

Conflicts of Interest/Competing Interests

Over the past 5 years, Stéphane Jouneau declares conflicts of interest in relation to the subject of this article as he was a primary investigator in IMPALA trials and a co-author of the international guidelines on PAP management, and outside of patients with aPAP, Stéphane Jouneau has declared links with the following pharmaceutical companies (or similar): clinical research: AIRB, Astra Zeneca, Biogen, BMS, Boehringer Ingelheim, Galacto, GSK, Pliant Therapeutics, Roche, and United Therapeutics; advisory boards, consultancy: AIRB, Boehringer Ingelheim, GSK, and Sanofi; courses, training: AIRB, Astra Zeneca, BMS, Boehringer Ingelheim, Chiesi, GSK, LVL, Novartis, Pfizer, Roche, and Sanofi; and research grants: AIRB and Boehringer Ingelheim. Over the past 5 years, Pierre Chauvin reports no conflicts of interest related to the subject of this article. However, he declares affiliations with the following pharmaceutical companies (or related entities): Boehringer Ingelheim, GSK, MSD, Pfizer, Takeda France, Fujifilm France, and Novatech SA. Over the past 5 years, Mathieu Lederlin reports no conflicts of interest related to the subject of this article. However, he declares affiliations with the following pharmaceutical companies (or related entities): Boehringer Ingelheim, Sanofi, Sophia Genetics, and AstraZeneca. Over the past 5 years, Benoît Painvin reports no conflicts of interest related to the subject of this article or outside of it. Over the past 5 years, Mallorie Kerjouan reports no conflicts of interest related to the subject of this article. However, she declares affiliations with the following pharmaceutical companies (or related entities): Boehringer Ingelheim, Grifols, CSL Behring, LFB, SOS, and Oxygène.

Ethics Approval

Not applicable.

Consent to Participate

Not applicable.

Consent for Publication

Not applicable.

Availability of Data and Material

Not applicable.

Code Availability

Not applicable.

Authors’ Contributions

SJ has written the first draft of the manuscript, all authors have read and amended the manuscript. All authors have read and approved the final version of the manuscript, and agree to be accountable for the work.