Whole lung lavage in alveolar proteinosis: manual clapping versus mechanical chest percussion

Abstract

Alveolar proteinosis is an uncommon lung disease presenting in primary or secondary forms, characterised by surfactant derived proteinous material accumulation within the lungs. The most effective treatment remains whole lung lavage under general anaesthesia. We have recently performed whole lung lavage in a 46-year-old patient with alveolar proteinosis who presented with severe dyspnoea and hypoxia. During the left lung lavage, outwards flow was enhanced at random either by manual clapping or by mechanical chest percussion with a vest airway clearance system. The protein and surfactant protein A concentrations in the 13 successive samples of the left lavage solution showed an exponential decline, not different between manual clapping and chest mechanical percussion. The average concentration of surfactant protein was not different between manual clapping and chest percussion. We conclude that in alveolar proteinosis, manual clapping replacement by mechanical chest percussion during whole lung lavage merits further evaluation.

BACKGROUND

The aim of the study was to verify whether a labour intensive procedure could be performed with less stringent “human” work without any loss in efficacy.

CASE REPORT

A 46-year-old male smoker, previously in good health, presented with epilepsy secondary to multiple cerebral micro-abscesses of Nocardia asteroides as confirmed by cerebral biopsy. He received trimethoprim-sulfamethoxazole as treatment. Afterwards the patient’s condition deteriorated rapidly with dyspnoea at rest, productive cough and severe hypoxia. Chest examination revealed bilateral extensive pulmonary crackles. The chest x ray and computed tomography (CT) scan (fig 1A,C) showed a diffuse, bilateral, micro-nodular pattern infiltration, predominating in the central areas with geographic distribution and a “crazy paving” pattern with association of diffuse ground glass areas. A surgical lung biopsy confirmed the diagnosis of alveolar proteinosis. Two successive whole lung lavage procedures were performed at 1 month interval, first in the right, then in the left lung, under general anesthesia with a double lumen tracheal tube.

Figure 1. (A, upper left) Chest x ray, before lavage, showing a diffuse, bilateral, alveolar infiltrate with a nodular pattern, with a central predominant distribution.

(B, upper right) Chest x ray, after unilateral right lung lavage, showing attenuation of the right lung infiltrates. (C, lower left) High resolution computed tomography section of the chest passing below the main carina revealing bilateral air space disease consisting of diffuse infiltrates with a geographic distribution showing a “crazy paving” pattern, with alternating diffuse ground glass areas, and thickening of the inter-lobular septa. (D, lower right) High resolution computed tomography of the chest, after unilateral right lung lavage confirmed the disappearance of the alveolar infiltrates in the right lung.

Lung lavage

As described by Ramirez et al,^1,2 the patient received successive 1 litre aliquots of saline solution while he was ventilated on 60% oxygen through the other lung. The first time, during the voiding phase, after each 1 litre infusion, the physiotherapist performed vigorous manual percussion of the chest at about 4 Hz.²

The second lavage occurred 1 month later, on the left lung. This time, the voiding phase was modified. We randomly assigned the mechanical aid to either manual clapping or mechanical chest percussion with a “Vest airway clearance system”, a bag that is inflated and deflated at rapid frequencies with an air pump,³ applied over the left chest wall at a frequency of 10 Hz with an amplitude of 3 arbitrary units.

During both lavages, the procedure was repeated until the fluid recovered was clear.

In order to assess whether manual clapping or mechanical chest percussion were similar in efficacy, during the second lavage procedure we measured in each sample of the recovered fluid total protein and surfactant protein A concentrations (for the methods, see below).

For the right lung lavage, 16 litres of saline were infused then recovered before the liquid became clear. For the left lung, 13 litres were needed. The average volume recovered during the physiotherapist assisted aliquots was 970 ml, whereas during vest assisted aliquots it was 1000 ml. Figure 1B,D shows the chest x ray and CT scan 3 weeks after the lavage of the right lung. The large difference between the two lungs illustrates the beneficial effect of the lavage of the right lung. Table 1 shows the evolution of clinical and physiological parameters. There was a rapid change in arterial blood gas values, accompanied by an increase in lung diffusing capacity (despite a transient decrease of the forced expiratory volume in 1 s (FEV₁) and vital capacity) and a significant lengthening of the 6 min walking distance. The patient improved his ability to perform the activities of daily living.

Table 1. Lung function, 6 min walking distance and arterial gas blood values throughout the evolution.

	Before right lung lavage	After right lung lavage	Before left lung lavage	After left lung lavage	3 weeks later	4 weeks later
Date	29 Nov 2006	1 Dec 2006	20 Dec 2006	22 Dec 2006	15 Jan 2007	19 Feb 2007
Vital capacity (litres) (%)	2.95 (57)	2.94 (57)	3.01 (59)	2.3 (44)	NA	3.23 (62)
FEV1 (litres) (%)	2.20 (57)	2.02 (52)	2.02 (52)	1.36 (35)	NA	2.29 (59)
Total lung capacity (litres) (%)	4.22 (62)	4.19 (61)	4.31 (63)	4.45 (65)	NA	4.63 (67)
DlCO (ml/mm Hg/min) (%)	9.3 (30)	12.6 (41)	11.3 (37)	12.5 (41)	NA	12.6 (41)
6 minutes walking distance (m)	470	620	580	NA	NA	NA
Pao2 (mm Hg)	52	59	70	63	96	NA
Paco2 (mm Hg)	29	30	32	33	30	NA
pH	7.43	7.42	7.42	7.40	7.42	NA

DlCO, lung diffusing capacity for CO; NA, non-available.

Values in parentheses are percentage of predicted values.

The total protein concentration in the 13 successive samples of the left lavage solution showed an exponential decline, as lavage proceeded, which was not different between manual clapping and mechanical chest percussion (fig 2). The mean concentration (all samples taken together) of surfactant protein A was not significantly different by unpaired Student t test between manual clapping and chest percussion (mean (SD) 1.30 (0.08) vs 1.32 (0.11) μg/mg protein)

Figure 2. (A, top) Concentration of proteins (mg/ml) during the second lavage (left lung) in the successive samples, from 01 (farthest left) to 13 (farthest right), after physiotherapist manual percussion (dark bars) and after mechanical procedure (light bars).

(B, bottom) Photograph of the 13 successive samples of the recovery fluid during lavage of the left lung. The first sample is on the left and the last sample on the right; note the decreasing level of milk-like fluid from left to right.

INVESTIGATIONS

Biochemical assays

Total protein concentration was assessed by the bicinchoninic acid method (Pierce, Rockford, Illinois, USA) following the manufacturer’s protocol. Surfactant protein A (SP-A) was assayed by direct ELISA; 96-well microplates were coated for 2 h at 37°C with lavage samples (diluted 1/200 in carbonate bicarbonate buffer, pH 9.6) or with human SP-A standard (gift from Dr Joanna Floros). After washes and blocking of non-specific sites with 1%, w/v BSA (Sigma Aldrich, St Louis, Missouri, USA), plates were incubated with rabbit antihuman SP-A IgG antibody (2 μg/ml, gift from Dr Joanna Floros) and, after washes, with HRP-conjugated anti-rabbit IgG (A-0545, Sigma Aldrich). Plates were then developed by incubating with hydrogen peroxide and tetramethylbenzidine (Pierce) and optical density was read at 450 nm using a plate spectrophotometer.

DISCUSSION

Whole lung lavage is a labour intensive procedure. As classically described by Ramirez et al¹ it requires the cooperation of experienced anaesthetists and pneumologists, and it needs the presence of an enduring, experienced chest physiotherapist. Our aim was to verify whether the latter could be “replaced” with mechanical assistance. Our results show that the non-inferiority hypothesis merits further evaluation: the Vest Airway Clearance System achieved similar results as compared to the physiotherapist (that is, similar protein concentration in the outward fluid as shown in fig 2A and similar surfactant protein A concentration in fluids recovered after manual clapping or mechanical percussion). This may render therapeutic management of alveolar proteinosis a little less demanding for the treating teams. We observed a greater immediate transient decrease in vital capacity and especially in the FEV₁ after the second whole lung lavage (where mechanical percussion was partly applied) than after the first lavage (where only manual clapping was used). Nevertheless, the lung diffusing capacity still increased, and the clinical response of the patient was as good after the second procedure as after the first. Further evaluation should closely monitor lung function after mechanical chest percussion.

We conclude that mechanical chest percussion, instead of manual clapping, merits further assessment in whole lung lavage for alveolar proteinosis.

LEARNING POINTS

• Whole lung lavage with vigorous clapping remains the cornerstone of symptomatic treatment in alveolar proteinosis.

• In many patients, repetitive procedures have to be performed.

• Manual clapping can be replaced by mechanical chest percussion.