Will inhalational GM-CSF replace whole lung lavage as a treatment for autoimmune pulmonary alveolar proteinosis? Many pole positions, not yet the final winner

Autoimmune pulmonary alveolar proteinosis (aPAP) is associated with autoantibodies against granulocyte–macrophage colony-stimulating factor (GM-CSF) in serum, causing the disruption of GM-CSF signal and the functional failure of alveolar macrophages and neutrophils [1]. As a result, surfactant phospholipids and proteins accumulate in the alveoli, leading to respiratory insufficiency [1]. In addition, aPAP patients are prone to opportunistic respiratory infections, mainly related to neutrophil dysfunction [2]. Whole lung lavage (WLL), which mechanically removes accumulated surfactant from alveoli, is still considered the standard of care for pulmonary alveolar proteinosis (PAP) syndrome, although it is time-consuming, expensive and fraught with potential risks to patients [3, 4]. Moreover, the WLL procedure is not yet standardised and can widely vary across centres [4].

Inhaled GM-CSF (iGM-CSF) is used worldwide for the treatment of aPAP, and although its use is still off-label [2], it is being used increasingly both inside and outside of clinical trials. The advantages of iGM-CSF are potentially related to its ability to restore alveolar macrophage and myeloid cell function, which translates into a clinical improvement and reinstatement of native immunity (figure 1) [1, 2, 5, 6].

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1

Short- and long-term effects of inhaled granulocyte–macrophage colony-stimulating factor on a) diffusing capacity of the lung for carbon monoxide (D_LCO), b) alveolar–arterial oxygen tension gradient (P_A-aO2) and c) Borg scale for dyspnoea assessment in autoimmune pulmonary alveolar proteinosis (aPAP). d) Chest high-resolution computed tomography (HRCT) of a 39-year-old female patient with aPAP, showing extensive crazy paving. e) Off-label treatment with inhaled sargramostim (250 μg daily) improved her clinical condition and HRCT findings, and no further whole lung lavage was needed. Figure 1a, b and c reproduced with permission from [13]. Figure 1d and e provided by the authors.

So far, 15 studies examining the effect of GM-CSF in adult aPAP patients have been published; 10 observational studies, 1 retrospective study, two randomised clinical trials and two case series [7–21]. In the early studies, GM-CSF was administered subcutaneously [7–10], whereas more recent ones have used iGM-CSF [11–21]; both routes of administration have demonstrated good safety and tolerability. In 2010, Tazawa et al. [11] published the first national, prospective, multicentre, phase II trial examining 35 aPAP patients with an arterial oxygen tension (P_aO2) of less than 75 mmHg that underwent sequentially high-dose recombinant human (rh) GM-CSF therapy over 52 weeks. Individuals were excluded if they had received WLL in the previous 6 months before enrolment. The overall response rate was 62% at 6 months and was maintained in 83% of the patients at 1 year, without the need for WLL rescue treatment. No safety or tolerability issues were observed. In 2019 and 2020 the first randomised placebo-controlled trials were published. In the Japanese PAGE trial, inhaled rhGM-CSF sargramostim at a dose of 125 μg twice daily for 7 days, every other week for 24 weeks, or placebo was administered to 64 patients with mild to moderate aPAP and no WLL within the previous 6 months [14]. iGM-CSF was associated with a significant improvement of alveolar–arterial oxygen tension gradient (P_A-aO2) and a consistent effect in reducing ground-glass opacities on computed tomography [22]. In the most recent worldwide phase III trial IMPALA, patients with aPAP and a P_A-aO2 of 25 mmHg or more, were randomly assigned to receive the recombinant GM-CSF (molgramostim 300 μg once daily by inhalation), either continuously or intermittently (every other week), or matching placebo. The 24-week intervention period was followed by an open-label treatment extension period with alternate week treatment [17]. Again, patients with WLL in the 1 month prior to enrolment were excluded. This was the first trial to demonstrate that daily administration of i-GM-CSF resulted in greater improvements in pulmonary gas transfer and functional health status than placebo, with similar rates of adverse events. Despite these successful trials, several unresolved issues about iGM-CSF remain. Due to the different trial design and study population, we cannot draw any conclusion whether sargramostim or molgramostim is better in terms of efficacy in treating aPAP. Further, the optimal iGM-CSF administration protocol (continuous versus intermittent, and if intermittent, which timing) is still debated in the PAP community [23–25].

In this issue of the European Respiratory Journal, Campo et al. [26] present the results of an Italian clinical trial, sponsored by the Italian drugs approval agency AIFA, which assessed the efficacy of iGM-CSF in aPAP patients after receiving one WLL at baseline. This sequential treatment enabled a reduction in additional WLL and improvement of lung function over the study period. This prospective, randomised, single-centre phase II study was designed by Professor Maurizio Luisetti in 2009 and was aimed at recruiting patients with moderate to severe aPAP, defined by a resting P_aO2 of less than 60 mmHg, or if more than 60 mmHg then a resting peripheral blood oxygen saturation of less than 90% or a 5% or greater decline during exercise. All patients underwent bilateral WLL within the month before treatment and then randomised into iGM-CSF and no treatment control groups, and followed up for 30 months. Treatment administration was limited to 10 months and the adopted protocol for iGM-CSF administration was based on a rather complicated intermittent schedule (decided a long time before the publication of the new iGM-CSF trials). However, despite several methodological concerns, such as the open-label design and the small sample size, the study demonstrated that iGM-CSF was able to reduce the requirement for further WLL and to improve lung function, and was safe. Two aspects differentiate the present trial from the previous ones: first, the inclusion of patients with severe disease. It has been speculated in the past that PAP patients with severe disease seem to respond more markedly and quickly to iGM-CSF treatment than those with mild to moderate disease. It is somehow disappointing, that in the newest published trials, patients requiring WLL in the few months prior to screening have been excluded. The second peculiar aspect concerns the selection of “time to first rescue WLL” as the primary end-point instead of the change in P_A-aO2 between baseline and end of study. Although the primary end-point is not a validated one, the strategy adopted by the principal investigator highlights the “wisdom and avant garde thinking” of choosing parameters of paramount clinical importance for the patients, rather than laboratory or functional values. The encouraging efficacy and safety results increase the clinical significance of combining WLL and GM-CSF inhalation, but at the same time raises several intriguing questions about trial initial design, which is now 14 years old. The first one challenges the treatment duration of 10 months. As reported by the authors, vital capacity, diffusing capacity of the lung for carbon monoxide, ground glass opacity score and Short Form-36 General Health Score present no significant difference between treated and untreated patients at month 10, which is compatible with significant residual disease or late response. What would have happened if effective treatment was protracted till complete remission? On the other side, the time to further WLL procedure in PAP patients has been reported to be around 200 days, or 7 months, but with huge variability between patients [3]. Furthermore, as stated by the authors in the discussion, daily iGM-CSF administration with a higher dose would have probably resulted in a larger improvement (but the trial was designed in 2008). Another burning point is about why off-treatment patients who experienced deterioration were only assigned to WLL and not iGM-CSF as a rescue treatment. Why change the winning team?

Finally, why interrupt the only effective pharmacological approach potentially able to restore lungs integrity and not maintain a lowest effective dose until the end of follow up (30 months), according to the “as far as it takes” paradigm? Indeed, no collateral effects were observed in the short and in the long term, in line with previous studies demonstrating no significant stimulating activity on haematopoiesis.

After a careful evaluation of all the studies on iGM-CSF in aPAP patients, we might conclude that an administration protocol in different phases could be depicted as follows: 1) start iGM-CSF daily; 2) titrate after remission; 3) find the lowest effective dose to sustain remission; and 4) suspend treatment after a consolidated disease-free period. In patients that have remained off-treatment and that experience deterioration, re-administer iGM-CSF as initially. Could the adoption of the above reported corrective measures in future trials lead to the final approval of iGM-CSF as the standard of care of the disease, keeping aside WLL as a “rescue” option only?

The present study, based on its long-term, prospective, randomised design and its positive results in terms of reduction of the need for “rescue” WLL and improvement of lung function and gas exchange in patients with moderate to severe aPAP provides another “pole position” for sequential treatment with WLL and iGM-CSF in the race of aPAP treatment trials. In conclusion, we believe that the vision of Maurizio Luisetti and the effort of his co-workers in making this complex study possible in a rare, neglected disease like aPAP, will pave the way to identifying the best treatment option for aPAP (iGM-CSF alone or in combination), even beyond the results of the past and ongoing clinical trials.

Shareable PDF