European Respiratory Society

Inhaled all-trans retinoic acid in an individual with severe emphysema

M. Frankenberger, I. Heimbeck, W. Möller, S. Mamidi, G. Kaßner, K. Pukelsheim, M. Wjst, M. Neiswirth, P. Kroneberg, D. Lomas, D. Halsall, P. Iadarola, A. Fertl, K. Häußinger, L. Ziegler-Heitbrock

To the Editors:

A patient with extensive emphysema was treated with liposomal all-trans retinoic acid (ATRA) using a controlled inhalation manoeuvre. After a gradual increase from 0.1 mg to 3 mg ATRA per day, the patient was treated at a level of 1 mg·day−1 for 4 days per week. Over the 6-month observation period there was an improvement of forced expiratory volume in 1 s (FEV1) from 16% to 21% predicted and an increase from 400 m to 615 m in the 6-min walking test (6MWT), although there was no apparent change in lung morphology with high-resolution computed tomography (HRCT). Biochemical analysis demonstrated a decrease in exhaled isoprostane and a decrease in serum interleukin (IL)-13. In addition, desmosine excretion decreased from 115 μg·g−1 creatinine to <35 μg·g−1 creatinine (normal range 25±4 μg·g−1). The data suggest that inhaled ATRA reduces inflammation and lung destruction processes and may be of benefit in selected cases of emphysema.

The patient was a 45-yr-old male never-smoker of Caucasian origin with no family history of lung disease. He had worked in the wood industry and then as an electrician without any known asbestos exposure for 7 yrs until diagnosis. He first developed dyspnoea in 2002 and HRCT analysis revealed extensive basal centrilobular emphysema. Standard inhalation therapy was Spiriva® (tiotropium 18 μg q.d.; Boehringer Ingelheim Pharma, Ingelheim, Germany) and Symbicort® (budesonid 160 μg plus formoterol 4.5 μg b.i.d.; AstraZeneca, London, UK). The patient was admitted to hospital several times with Haemophilus influenzae airway infection and, as of summer 2007, he was on long-term oxygen therapy at 2–3 L·min−1 for 24 h per day. The α1-antitrypsin level was normal at 171 mg·dL−1 and sequencing of the SERPINA1 (AAT) gene (exons plus flanking sequence) showed the patient to be homozygous for the common M1V variant with no other sequence variation.

Retinoic acid (RA) acts via an retinoid X receptor–RA receptor-α nuclear receptor complex. RA has been demonstrated to promote alveolarisation in a rat model 1, but other reports in rats and mice failed to demonstrate such an effect 25. Furthermore, ATRA can improve the protease/anti-protease balance 6, 7 and has anti-inflammatory activity 8. Oral therapy with ATRA in chronic obstructive pulmonary disease (COPD) has been reported and there is evidence of some improvement 9, 10. We reasoned that topical application to the lung might be superior to systemic oral therapy.

Therefore, the patient was considered for compassionate use of liposomal ATRA applied by controlled inhalation. For this di-oleoyl-phosphatidyl serine (OOPS) and 1-palmitoyl-2-oleyl-phosphatidyl choline (POPC) (Mandapro AG, Ettingen, Switzerland) were dissolved at an OOPS/POPC ratio of 0.43 in chloroform 6. ATRA (#700121-0005; Fagron, Barsbüttel, Germany) was dissolved in 96% ethanol at 2 mg ATRA mL−1. On a weekly basis, 175 mg of lipids was admixed with 7 mg of ATRA, and liposomes were generated and extruded to ≤0.4 μm by passage through polycarbonate filters (Nuclepore #110607; Costar, Bodenheim, Germany) in an extruder device (B002; Lipex Biomembranes, Vancouver, BC, Canada). Resultant unilamellar liposomes were passed through a 0.2-μm filter and stored at 4°C for use within 2 weeks.

After having given informed consent, the patient inhaled ATRA in 1- to 3-mL volumes of liposomal preparations. Inhalation was controlled using the AKITA2 APIXNEB® device (Activaero, Gemünden/Wohra, Germany) with 500 mL of the liposomal ATRA aerosol followed by 100 mL of air only. The drug was administered three times a week (Tuesday, Wednesday, Thursday) starting with a dose of 0.1 mg, with a weekly increase to 0.3 mg, 1 mg and finally 3 mg of ATRA.

There were no side-effects reported by the patient, and physical and laboratory examination did not show any evidence of adverse effects. After 4 weeks of incremental therapy the patient was continued on ATRA at 1 mg, 4 days a week (Tuesday, Wednesday, Thursday, Friday) for 5 months. From then on, the patient was continued on ATRA therapy without further follow-up and the restrictions on additional therapies and supportive measures (e.g. exercise training) were lifted. This report covers 6 months of controlled ATRA therapy.

After several weeks of therapy, the patient noted an improved performance in daily life and he was able to spend several hours without oxygen supplementation on weekends. Analysis of lung function before therapy and at the end of the study period showed an improvement in FEV1 and blood gases. In addition, there was a strong increase in the standard 6MWT (table 1).

View this table:
Table 1—

Lung function and laboratory parameters

Although better supportive care during the study period may have added to the clinical improvement, we found clear biochemical changes that support a specific action of liposomal ATRA. This includes a decrease in plasma IL-13, neutrophil elastase in sputum and desmosine in urine (table 1). In addition, sputum eosinophils were increased to 12% before ATRA treatment without any evidence of asthma in lung function testing (table 1). Eosinophils decreased to 6%, a phenomenon that may be due to an ATRA-induced shift to neutrophils 11. 8-isoprostane levels determined in exhaled breath condensate showed a substantial decrease from 14.6 pg·mL−1 before the study to 4.3 pg·mL−1 at the end of the study. Also, IL-13 in plasma, as detected by multiplex ELISA, was high before therapy and decreased over time (table 1). There was no clear pattern in the matrix metalloproteinase-9/tissue inhibitor of metalloproteinase-1 ratio in patient plasma and sputum (data not shown).

Elastin degradation by proteases is an important element of lung destruction in emphysema and this can be monitored via the levels of the amino acid desmosine, which is released upon digestion and excreted in urine. Desmosine was measured via capillary electrophoresis 12. The levels of this degradation product showed a gradual decrease and were near the normal range at 6 months (table 1).

Regarding the mechanisms of action of liposomal ATRA, we did not have tools with the appropriate sensitivity to demonstrate an increase in alveoli in this clinical setting. However, we found clear evidence for an anti-inflammatory action of liposomal ATRA, with a decrease of 8-isoprostane in exhaled breath condensate and a decrease of the IL-13 plasma levels. There is conflicting evidence of the role of IL-13 in typical COPD patients 13, 14, but in the present case IL-13 was substantially increased and responded to liposomal ATRA. When looking at sputum fluid there was a decrease in neutrophil elastase. This decrease may explain, in part, a reduced degradation of elastin as evidenced by a decrease in the desmosine excretion. The latter end-point is obviously crucial to the course of the disease since it indicates a blockade of further lung destruction.

Statement of interest

Statements of interest for M. Frankenberger, D. Lomas and L. Ziegler-Heitbrock can be found at


We would like to thank P. Fankhauser (Mandapro, Ettingen, Switzerland) for generous provision of lipids.