Positive

Presen t

Fig. 22.4  Algorithm used for diagnosis of PAP syndrome. The presence of PAP is suspected based on a compatible history, typical radiologicndings, and bronchoalveolar lavage cytology ndings. GM-CSF autoantibody test should be performed initially when PAP is suspected: a positive test con rms the diagnosis of autoimmune PAP. Patients with a negative GM-CSF autoantibody test that have a disease known to cause PAP are diagnosed with secondary PAP. Those with a negative GM-CSF autoantibody test and no underlying PAP-causing disease should undergo a blood-based GM-CSF signaling test and serum GM-CSF test: individu-

als with positive tests should undergo further tests for GM-CSF receptor gene (CSF2RA or CSF2RB) mutations to identify hereditary PAP. Patients with a negative GM-CSF signaling test should undergo further tests for other gene mutations (e.g., in SFTPB, SFTPC, ABCA3, or NKX2.1) to diagnose surfactant production disorders. If no PAP-causing disease can be found, a transbronchial or surgical lung biopsy for lung parenchymal histopathological examination may be needed. (Adapted from Trapnell et al. Pulmonary alveolar proteinosis. Nat Rev Dis Primers. 2019 Mar 7;5(1):16. doi: 10.1038/s41572-019-0066-3. PMID: 30846703)

Treatment

Shortly after PAP was rst reported, a number of therapeutic strategies were tested in case studies, including administration of antibiotics, corticosteroids, and acetylcysteine among others [119]. Whole-lung lavage (WLL) emerged early and has remained the standard treatment. The serendipitous discovery of a PAP in Csf2KO mice and the subsequent discovery of GM-CSF neutralizing antibodies in patients with (then) idiopathic PAP shifted the focus toward evaluation of

GM-CSF augmentation therapy and others, including plasmapheresis and anti-B lymphocyte immunotherapy. We focus on WLL and GM-CSF augmentation therapy.

Whole-Lung Lavage

WLL is widely considered to be appropriate rst-line therapy of autoimmune PAP. Ramirez introduced the use of pulmonary saline instillation as therapy of PAP in the early

1960s with a procedure in which a transtracheal catheter was used to drip saline into the trachea multiple times per day for several weeks to induce cough-mediated surfactant clearance. However, the approach was not well tolerated or accepted into general practice. Subsequently, the procedure for instilling saline into the lungs underwent re nements, such as use of a double-lumen endotracheal tube and use of large volumes of saline that allowed the lungs to be more effectively and safely removed in a single session—a procedure now referred to as WLL. Briefy, saline is administered via a double-lumen endotracheal tube under general anesthesia, thereby allowing one lung to be “washed” to physically remove the excess surfactant while the other lung is mechanically ventilated [120–122]. Notwithstanding improvements, WLL is not been standardized across institutions with respect to the method, indications for its use, evaluation of treatment responses, or the timing of repeated procedures.

The WLL procedure is dependent on operator experience and, ideally, should be performed by an experienced team including an interventional pulmonologist or surgeon, anesthesiologist, a nurse, and a respiratory therapist [123]. Briefy, general anesthesia is induced and the patient is intubated with a double-lumen endotracheal tube to isolate each lung, and positioned for the procedure. Various body positions are used at different centers, including treated side up or down, supine, lateral decubitus; however, little supporting evidence has been reporting to identify the optimal approach [124]. The non-treated lung is ventilated with 100% oxygen to optimize oxygenation. Lung isolation is con rmed and then saline pre-warmed to 37 °C is instilled under a small (30 cm) hydrostatic head in infused in volumes of 500–1000 mL aliquots and drained under gravity via a closed ­system [22]. Chest percussion is used to loosen the surfactant sediment and emulsify it with the saline— either by manual percussion or using a wrap-around, pulsating vest [125, 126]. Each cycle of saline infusion, chest percussion, and drainage typically required 3–5 mins to complete (or longer if asthma is present). This process is repeated until the milky/turbid effuent clears, which can require up to 50 L of saline per lung in adults although most patients clear with about 30 L and smaller volumes in children. Patient monitoring during the procedure includes continuous measurement of peripheral blood oxygen saturation (SpO2) and vital signs, serial measurement of Pa02, tracking infusion and effuent volumes, observation for leakage of saline into the non-treated (ventilated) lung, and monitoring for adverse events, such as leakage of saline into the pleural space. Upon completion of infusion/drainage cycles, the lungs are examined bronchoscopically and residual saline is aspirated, ventilation with 100% oxygen is resumed, and the double-lumen tube is replaced with a single-lumen tube until extubation can be performed safely. In high risk adults, the use of extracorporeal membrane oxygen [127], hyperbaric conditions [127, 128], or bron-

choscopic segmental and lobar lavage can be used to reduce intraoperative risk [129].

The indications for WLL therapy are not standardized among institutions but exertional dyspnea-limiting physical activity is a common primary indication [125, 126]. WLL should be considered in patients with declining lung function, PaO2 or SpO2, a shunt fraction >10–12%, or radiographic evidence of disease progression [122]. The most severely affected lung or lung segment is usually treated rst and the less affected lung is treated several days to weeks later. Active bacterial lung infection, sepsis, and shock are contraindications for WLL [125, 126].

While no randomized controlled trials evaluating the safety and ef cacy of WLL have been reported in the medical literature, WLL it is widely regarded as capable of improving patients’ symptoms, radiographic abnormalities, and oxygenation [130, 131]. One study described the WLL outcomes for 231 individuals with PAP, reporting an overall 5-year survival of 94% with lavage versus 85% without lavage [2]. This study reported WLL was typically performed within 5 years of diagnosis (70% of patients) and the median number of lavages required was two per patient. In one subgroup analysis (n = 55), only 20% of patients remained recurrence free at 3 years and the median duration of bene t was 15 months. A demonstrable improvement in arterial PaO2 of 20.1 mmHg was noted among 41 patients for whom data were available with less impressive improvements in other pulmonary physiology parameters. Overall, WLL was an effective intervention with a response in over 95% of patients with pulmonarybrosis noted with greater prevalence in the remaining 5% who failed to respond. Treatment-­related improvements are often noted within hours or days after the procedure. WLL is generally well tolerated and safe but nonetheless carries a risk of several uncommon complications, including persistent hypoxemia, pneumonia, sepsis, hydropneumothorax, and acute respiratory distress syndrome.

Subcutaneous GM-CSF

In 1996 (shortly after the discovery of PAP in GM-CSF-­ de cient mice), recombinant GM-CSF therapy was reported in a patient with (then) idiopathic PAP who experienced a treatment-related reduction in PAP-related symptoms and improvement in PaO2 [132]. Subsequently two small open-­ label studies evaluated subcutaneous GM-CSF administered in autoimmune PAP in escalating doses over a period of 3 months or 6–12 months, respectively. The overall treatment response rates (de ned as a 10 mmHg improvement in room air A-aD02) in the two studies were 43% and 48%, respectively [93, 133]. Improvement was also observed in DLCO, radiographic abnormalities, and the distance walked in a 6MWT (in the second study). Subsequent case studies and small series reported similar ndings with objective

improvement noted in lung disease severity in approximately 50% of patients. The treatment result varied depending on the dose and duration of administration. Injection site irritation occurred in 85% of patients in one study [133], but there was no signi cant treatment-associated adverse events reported in any of the referenced studies.

Inhaled GM-CSF

In 2005, aerosolized GM-CSF therapy was rst reported in three patients with autoimmune PAP in whom treatment-­ related improvements were observed in A-aD02, and microscopic abnormalities seen in BAL cytology [89, 134, 135]. Subsequently, other small studies reported similar positive results [136]. In a 64-week retrospective study in which 12 autoimmune PAP patients received inhaled GM-CSF in escalating doses (250–500 μg, twice daily every other week), 11 had improvements in AaD02 and DLCO [137]. In a subsequent prospective open-label study, 39 patients with unremitting or progressive autoimmune PAP were administered inhaled GM-CSF in two phases: a high-dose induction phase (250 μg daily on days 1–8, every 2 weeks for 3 months) followed (in 35 of the 39 patients) by a low-dose maintenance phase (125 μg, days 1–4, every 2 weeks for 3 months) [138]. Of the 35 patients receiving both induction and maintenance therapy, 24 (62%) experienced improvement in A-aD02 (the primary endpoint) and the treatment was found to be safe and well tolerated. A follow-up study found 23 of 35 patients had a durable treatment response [139]. A randomized, placebo controlled open-label study of inhaled GM-CSF therapy (150 μg twice daily, alternate weeks, 24 weeks) in 36 patients with mild–moderate autoimmune PAP conducted in China reported improvement in DLCO and St George Respiratory Questionnaire score but not in A-aD02, radiographic abnormalities, or WLL requirement [140].

Recently, two randomized, placebo-controlled, double-­ blind studies have evaluated the safety and ef cacy of inhaled GM-CSF in autoimmune PAP, the PAGE, and IMPALA trials. The PAGE trial was conducted in Japan in 64 patients with mild–moderate disease (PaO2 <75 mmHg and >50 mmHg) [141]. The blinded treatment group (GM-CSF 125 μg twice daily on alternate weeks, for 6 months) demonstrated improvement in AaD02, radiographic lung densitometry scores (based on chest CT scan evaluations), and reduction in levels of PAP biomarkers in the treatment group compared to the control group but no improvement in clinical or patient-reported outcomes. [141] No safety concerns were identi ed. The IMPALA study evaluated 138 autoimmune PAP patients and compared three intervention groups during a 24-week blinded treatment period: (1) continuous GM-CSF (300 μg daily), (2) intermittent GM-CSF (300 μg daily on alternate weeks with placebo on alternate weeks to

maintain the blind), and (3) placebo (inhaled daily). [142] Compared to placebo, signi cant improvement was observed in the continuous GM-CSF group as measured by change in AaD02, DLCO, radiographic ground-glass opaci cation score (based on chest CT scan evaluations), Saint George’s Respiratory Questionnaire Total Score, and PAP biomarkers. Further improvements were seen in an open-label treatment extension period of 48 or 72 weeks.

Inhaled GM-CSF was well tolerated and without any treatment-related serious adverse events. This study demonstrated inhaled GM-CSF resulted in improvement in clinical, radiographic, physiologic measures, quality of life, and was safe and well tolerated, and that continuous administration was more effective than daily administration on alternating weeks [142]. A subsequent study, IMPALA-2, was designed as a global phase 3 trial to evaluated the long-term ef cacy and safety of continuous daily inhaled GM-CSF therapy in patients with autoimmune PAP is currently underway. [22].

A small randomized open-label study in Italy reported marked reduction in the requirement for subsequent WLL therapy when WLL was combined with inhaled GM-CSF in 18 patients with autoimmune PAP [143]. Several case reports describe improved treatment responses when WLL is performed prior to administration of inhaled GM-CSF [144, 145].

In summary, results from numerous case reports, small patient series, and two randomized, placebo-controlled, double-­blind trials have repeatedly demonstrated the safety and ef cacy of inhaled GM-CSF, while still off-label, is a promising pharmacotherapeutic of autoimmune PAP. Treatment ef cacy is greater when administered by aerosol compared to subcutaneously, and when administered continuously compared to daily on alternating weeks. A starting dosing of 250–300 μg/day appears to be effective in half–two- thirds of patients and no safety concerns have been identi ed in any study. Further research is needed to determine if a dosedependent response exists as suggested by anecdotal clinical observations, and to determine the optimal dose, duration of treatment required for a maximal treatment response, and whether an induction–maintenance treatment strategy will be useful to maintain the treatment response [22].

Other Approaches

As previously mentioned, a number of other treatment approaches have been tested in small numbers of autoimmune PAP patients. The observation that GM-CSF was required for normal cholesterol metabolism by alveolar macrophages led to studies targeting cholesterol homeostasis.

The PPARγ agonist (pioglitazone) showed promising results in a mice studies [30] and human case reports [146] with results pending from a phase I/II trial of pioglitazone in autoimmune PAP. Furthermore, administration of statins