Copyright #ERS Journals Ltd 2004

Eur Respir J 2004; 23: 435–439 European Respiratory Journal

DOI: 10.1183/09031936.04.00084904 ISSN 0903-1936
Printed in UK – all rights reserved

A short-term comparison of two methods of sputum expectoration in

cystic fibrosis

K. Chatham, A.A. Ionescu, L.S. Nixon, D.J. Shale

A short-term comparison of two methods of sputum expectoration in cystic fibrosis. Dept of Physiotherapy and Section of Respira-
K. Chatham, A.A. Ionescu, L.S. Nixon, D.J. Shale. #ERS Journals Ltd 2004. tory and Communicable Diseases, University
ABSTRACT: The aim of this study was to determine whether repeated maximum of Wales College of Medicine, Academic
Centre, Llandough Hospital, Cardiff and
inspiratory vital capacity manoeuvres against a fixed resistance increased effective Vale NHS Hospital Trust, Penarth, UK.
short-term sputum clearance in adults with cystic fibrosis (CF).
Twenty adults with CF were randomised to receive, on alternate days, either Correspondence: D.J. Shale, Section of
standardised physiotherapy (SP) for 30 min, comprising postural drainage and the Respiratory and Communicable Diseases,
active cycle of breathing technique, or a series of resistive inspiratory manoeuvres University of Wales College of Medicine,
(RIM) at 80% of their maximum sustained inspiratory pressure developed between Academic Centre, Cardiff and Vale NHS
residual volume and total lung capacity during the first 4 days of the treatment of an Trust, Llandough Hospital, Penarth, Vale of
exacerbation of respiratory symptoms. Expectorated sputum was collected during and Glamorgan, CF64 2XX, UK.
for 30 min after each treatment and weighed. Total protein, immunoreactive interleukin Fax: 44 2920716416
E-mail: [email protected]
(IL)-8 and human neutrophil elastase (HNE) concentrations, and the amount of each
component expectorated, were determined. Keywords: Cystic fibrosis, physiotherapy, resis-
Compared with SP, RIM increased sputum weight two-fold, independent of tive inspiratory manoeuvres
treatment order or day. The concentrations of protein, IL-8 and HNE in sputum
were similar for both treatments, while the quantity expectorated was greater with RIM Received: July 22 2003
treatment. Accepted after revision: November 26 2003
In conclusion, short-term resistive inspiratory manoeuvres treatment was more effec-
tive at clearing sputum and inflammatory mediators than standardised physiotherapy. A.A. Ionescu and L.S. Nixon were supported
by CF Trust UK project grants. Other support
Eur Respir J 2004; 23: 435–439.
was from the Astra Foundation UK and
GlaxoSmithKline UK.

In cystic fibrosis (CF) airways secretions are thick, inflammatory response. Traditionally, physiotherapy has
tenacious and difficult to clear from the respiratory system, consisted of postural drainage with additional percussive,
particularly during periods of exacerbation of respiratory vibratory or expiratory manoeuvres to mobilise sputum from
symptoms. Abnormalities of airway secretions are due to the peripheral airways to larger central airways for clearance by
effects of altered CF transmembrane receptor function, and coughing. However, there is no consensus as to which
include alterations in ion transport and hypersecretion of approach is the most effective, although there are advocates
modified mucus [1–3]. Secondary to this are the effects of of different therapeutic regimens using various techniques and
chronic infection and the host inflammatory response, which devices to attain effective expectoration [13–18]. A meta-
leads to airways secretions rich in viable bacteria and their analysis of 35 studies concluded that standardised physiother-
exoproducts, host-derived neutrophils, plasma proteins, apy (SP) enhanced sputum clearance and improved the forced
deoxyribonucleic acid, tissue damaging enzymes and pro- expiratory volume in one second (FEV1), and that additional
inflammatory cytokines [1, 4, 5]. This process occurs from methods added little to SP [15].
early in life and becomes continuous in the majority of There is a growing population of adults with CF, many of
patients, leading to chronic sputum production [6]. Progres- who find it difficult to adhere to physiotherapy regimens, with
sive parenchymal lung injury leads to airway instability, reported levels of only 40–50% adherence [19]. This occurs for
promoting hyperinflation, dynamic collapse and impaired a variety of reasons, including not accepting the rationale for
clearance of sputum [3–6]. Interleukin (IL)-8, a major continued treatment and the problems of fitting treatment in
neutrophil chemotactic cytokine, is increased in sputum and with domestic or workplace pressures [20]. Treatments which
bronchoalveolar lavage obtained from patients with CF, and clear secretions with a minimum of disruption to lifestyle, but
the concentration is related to clinical severity and lung which are as effective as traditional forms of physiotherapy,
function [7–9]. The sputum levels of IL-8 and neutrophil may have a positive influence upon disease status due to an
elastase have been used as indicators of local inflammation in increased compliance with such treatments, although there is
CF, asthma and COPD [8–11]. little evidence for this at present [15].
Chest physiotherapy is an integral component of the treat- The current authors have previously described the use of a
ment plan of patients with CF [12–15]. A variety of treatment fixed-load method for assessing inspiratory muscle function,
approaches have been developed with the common goal of which can also be used for inspiratory muscle training [21].
maximising sputum clearance, which it is believed will reduce Patients in such studies reported that repeated inspiratory
sputum retention, atelectasis and the duration of exposure to manoeuvres against a resistance, a Muller manoeuvre,
injurious oxidant and proteolytic agents produced by the host resulted in increased sputum expectoration. This effect may
436 K. CHATHAM ET AL.

be similar to that seen in exercise, which has been associated 300

with increased sputum clearance as a secondary effect [13, 15].
To study this observation and its potential use in the manage- 270
ment of patients with CF, the effect of resistive inspiratory 240
manoeuvres (RIM) against a fixed resistance was compared 210

Pressure cmH2O
with that of supervised respiratory SP, incorporating the
active cycle of breathing, on sputum clearance in adult 180
patients chronically infected with Pseudomonas aeruginosa. In 150
addition, the impact of these treatments on the clearance of
inflammatory mediators during the first 4 days of the 120
treatment of an exacerbation of respiratory symptoms was 90
determined.
60
30
Material and methods 0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Patients Time s

Fig. 1. – A sustained maximum inspiratory pressure curve.

A total of 20 adult patients (10 female) with proven CF
(sweat Naz and Cl- w70 mmol?L-1 and an appropriate geno-
carried out repeated inspiratory manoeuvres based on this
type), with chronic pulmonary infection with P. aeruginosa
template, which they attempted to match on each breath.
(defined as more than six isolations from sputum in the
Each SMIP was expressed as Joules [21, 22]. This phase
preceding year), were studied. They were studied when
consisted of a maximum of up to 36 inspiratory manoeuvres
presenting with an exacerbation of respiratory symptoms,
in groups of six efforts, as described in the test of incremental
defined as a combination of increased cough, shortness of
respiratory endurance [21, 22]. The patient had to achieve
breath, sputum production and a reduction in FEV1 w10% of
90% of the template on any given breath to continue with the
their usual value with bacteriological confirmation of a heavy
series. The duration of treatment varied between patients,
growth of P. aeruginosa on culture of sputum. All were
depending upon each subject9s time to the point of failure.
admitted to hospital for intravenous antibiotic treatment and
Hence, each group of six inspiratory efforts was characterised
gave written informed consent to be included in the study,
by a shorter rest interval between each inspiratory man-
which had Local Research Ethics Committee approval.
oeuvre. The rest intervals reduced from 1 min to 45, 30, 15, 10
Participants were randomly allocated to alternate day
and 5 s. A leak calibration constant was calculated, from flow
treatment with either the RT2 inspiratory resistance device
rate (Q), as follows:
(DeVilbiss Healthcare UK Ltd, Wollaston, UK) or phy-
pffiffiffi
siotherapy for 4 days, starting from the first day of treatment Q~3:226|10
6 | p ð1Þ
with antibiotics, coinciding with the day of admission. For
each patient, treatment started with one or the other where pressure (p) was expressed in N?m-2 and Q in m3?s-1.
technique. Hence, both treatments were applied twice with Power (P; in watts) was then calculated as follows:
the patient randomly allocated to receive the same treatment P~p|Q ð2Þ
on days 1 and 3 or days 2 and 4. All treatment sessions were
performed under supervision and at the same time of the day. The work per breath was derived from the power curve and
All usual medications were administered during the study expressed in Joules [22]. All patients were familiar with the
days; the inhaled and/or nebulised treatments were standar- RT2 device and the RIM protocol.
dised and administered before the study interventions and Any sputum produced during either treatment was
were the same on all study days. expectorated into a pre-weighed container. At the end of
the treatment session, expectorated sputum was collected into
the same container for a further 30 min. The containers were
coded and the sputum analysed by the laboratory staff
Interventions unaware of the treatment group to which the patient had been
assigned.
SP comprised a supervised session of 30 min including
three postural drainage positions with percussion adminis-
tered by the physiotherapist and use of the forced expiration
technique [16]. Periods of relaxed breathing and thoracic Measurements
expansion exercises were also used as described in the active
cycle of breathing [14]. Patients were already familiar with, Sputum weight was derived from the total weight minus the
and trained in the use of, the active cycle of breathing pre-use pot weight. The same balance (Ohaus Precision
technique and postural drainage with percussion, which Advanced; Ohaus, Cottenham, Cambridge, UK) was used for
comprised the SP sessions. all measurements. Sputum was refrigerated until processing,
The inspiratory resistance protocol involved the use of the which comprised dilution 1:5 weight/weight with saline and
RT2 hand-held manometer with a fixed leak of 2 mm and 30 min mixing on a slow rotating mixer to liquefy the sample,
attached to a lap top computer loaded with the software which was then centrifuged at 10,0006g for 30 min. The upper
programme of the RT2 device. The maximum inspiratory sol-phase was removed and stored at -70uC until analysed.
pressure generated during a full inspiratory vital capacity Total protein concentration was determined by a modifica-
from residual volume (RV) was determined three times for tion of the micro-Lowry method, using albumin as a standard
each patient. This was shown as a sustained maximum (P5656; Sigma, Dorset, UK) [5].
inspiratory pressure (MIP) (SMIP) curve on a computer Human neutrophil elastase (HNE) was determined by a
screen (fig. 1). The software re-set the maximum pressure double antibody sandwich enzyme-linked immunosorbent assay
curve on screen at 80% of the original and the patients then (ELISA) developed for neutrophil elastase-a1-antiproteinase
 SPUTUM EXPECTORATION IN CF 437

Table 1. – The forced expiratory volume in one second 20

(FEV1), maximum inspiratory pressure (MIP) and sustained
maximum inspiratory pressure (SMIP) at commencement of 18
treatment in patients undergoing standard physiotherapy (SP)
or resistive inspiratory manoeuvres (RIM) 16
14
SP days 1 & 3, RIM days 1 & 3,

Sputum weight g
RIM days 2 & 4 SP days 2 & 4 12
10
FEV1 % pred 48.3 (37.6–59.0) 59.7 (40.1–79.3)
MIP cmH2O 132.2 (111.8–153.0) 132.4 (97.4–167.4) 8
SMIP J 9.8 (6.8–12.8) 9.6 (6.1–13.1)
6
Data are presented as mean (95% confidence interval). 4
2
complex and modified by using a mouse anti-human neutrophil
elastase (Dako, Ely, UK) coating antibody [5]. 0
IL-8 was determined by an in-house double antibody RIM days 1&3 SP days 2&4 SP days 1&3 RIM days 2&4
sandwich ELISA.
The laboratory researcher was blind to the treatment Fig. 2. – The mean weight of sputum expectorated during resistive
administered to patients. inspiratory manoeuvres (RIM; p) treatment and standardised phy-
FEV1, forced vital capacity (FVC) and their ratio (FEV1/ siotherapy (SP; h).
FVC) were determined by spirometry and results were
expressed as % predicted [23]. Height, weight and skin-fold Expectorated sputum
thickness were determined, and fat free mass (FFM) was
calculated [21]. A low FFM was defined as being less than the The mean weight of sputum expectorated during RIM
lower 5th centile for local healthy subjects of matched age and treatment was greater than for the SP and was not affected by
sex [24]. the order in which the treatments were given (fig. 2). The
pooled mean weight of sputum expectorated during and
following SP was 4.9 g (3.31–6.56) compared with 10.0 g
Statistical analysis (7.65–12.37) for the RIM method.
The concentration of protein, IL-8 and HNE (per g of
sputum) was not different between the treatments or between
Data were not normally distributed and were log10
days, nor was it affected by the order in which the treatments
transformed for analysis by t-test. Data are presented as
were given. The absolute amount of each component cleared
geometric mean and 95% confidence intervals. A paired-
was greater with the RIM method (table 2, fig. 3).
sample t-test was used to compare the concentrations of
The mean difference between the absolute amounts of
protein, IL-8 and HNE between the two types of treatment
protein, IL-8 and HNE obtained after RIM and SP were
and between the two sessions of the same type of treatment
66.3 mg (32.5–100.2), 4.6 ng (-2.0–11.4) and 297.3 mg
(i.e. RIM or SP). Spearman9s rank correlation test and linear
(60.1–34.5), respectively.
regression were used to determine relationships between
The FFM, FEV1 and inspiratory muscle function (MIP
variables.
and SMIP) were related, (all pv0.001). No difference was
found between patients with a low or normal FFM for the
clearance of any sputum component studied. For the whole
Results group, FFM or FEV1 had no influence on the amount or
concentration of any sputum components produced during
Patients SP or RIM.

The FEV1, MIP and SMIP were similar in the patients

commencing treatment with SP or RIM (table 1). The Discussion
proportion of patients with a normal or low FFM was also
similar between the treatment groups (Chi-squared p=0.66). This short-term study, carried out during the first 4 days of
Within the whole group, there were 11 patients with a low and antibiotic treatment for worsening respiratory symptoms,
nine with a normal FFM. The mean number of resistive demonstrated that repeated inspiratory manoeuvres against
inspiratory manoeuvres achieved was 24.6 (19.7–29.4; range a fixed resistance, standardised by visual bio-feedback to
8–36). This reflected a treatment time between 10 and 40 min, produce single-breath sustained inspiratory pressures of 80%
including rest periods, as described in the protocol. The of maximum, increased the wet weight of sputum expecto-
number of resistive breaths was related to FEV1 (r=0.83, rated over a 60-min period compared with SP. The increased
pv0.01). quantity of expectorated sputum enhanced the clearance of

Table 2. – Clearance of protein, interleukin (IL)-8 and human neutrophil elastase (HNE) in patients undergoing standard
physiotherapy (SP) or resistive inspiratory manoeuvres (RIM)
Total protein mg IL-8 ng HNE mg

SP 41.6 (19.5–89.1) 1.02 (0.45–2.24) 138.0 (54.9–346.7)

RIM 97.7 (63.1–154.9)** 1.47 (1.22–1.78)** 338.8 (119.5–562.3)*

Data are presented as geometric means (95% confidence interval). *: pv0.05; **: pv0.01.
438 K. CHATHAM ET AL.

70 treatment produced greater clearance of total protein, a

general indicator of increased microvascular exudation and
# cellular breakdown in the airways and cause of increased
60
sputum viscosity. The increased clearance of HNE, a measure
50 of the neutrophil load in the airways, indicates the potential
for removal of injurious products. This may be added to by
40 ** the greater clearance of IL-8, a potent chemo-attractant for
Units

neutrophils, during and after the RIM treatments [30]. These

30 comments are speculative and the demonstration of any long-
term benefit from such effects requires further studies.
20 As the lung disease progresses, airways instability leading
**
*** to collapse may confound the effectiveness of standardised
10 physiotherapy and other techniques may be needed at
different stages of the disease to achieve effective sputum
0 expectoration. The resistive inspiratory manoeuvres method
Sputum g Protein mgx0.1 IL-8 mgx10 HNE mgx0.1 may also have the advantage that its settings of pressure and
time are defined by the impaired pulmonary dynamics of the
Fig. 3. – Clearance of protein, interleukin (IL)-8 and human neutro- patient on each occasion it is used. It would be possible to
phil elastase (HNE) in sputum from patients undergoing standard adapt the software used in this method to enable patients to
physiotherapy (SP; h) or resistive inspiratory manoeuvres (RIM; u). accurately breathe from a known lung volume, as required by
#
: pv0.05; **: pv0.01; ***: pv0.001. protocols such as the active cycle of breathing techniques or
autogenic drainage. Further studies are required to investigate
protein, IL-8 and immunoreactive HNE from the airways. such possibilities. At present, the authors have demonstrated
The design of the study, with random allocation to the order that a series of near maximal inspiratory efforts against a
of treatments, allowed the authors to demonstrate that RIM fixed resistance appears to be more effective, in single
increased the weight of expectorated sputum irrespective of treatment episodes, than standardised physiotherapy for
the order of treatments or the day of treatment. sputum for expectoration in patients in a respiratory
An objective of physiotherapy is to increase expiratory flow exacerbation, and it leads to enhanced clearance of inflam-
rates with resultant high shearing forces, facilitating sputum matory and injury mediators.
mobilisation and transfer to more central airways. Hence,
treatments usually incorporate techniques to avoid dynamic
compression or collapse of airways caused by raised trans-
thoracic pressure, and the resultant distal movement of the References
equal pressure point during forced expiration and coughing
[25–27]. The finding in this study that RIM led to increased 1. Davies PB, Drumm M, Konstan MW. Cystic fibrosis: state
expectoration of sputum requires explanation, although the of the art. Am J Respir Crit Care Med 1996; 154: 1229–1256.
underlying mechanism in this study was not examined. The 2. Smith JJ, Travis SM, Greenberg EP, Welsh MJ. Cystic
increased weight of expectorated sputum suggests the possible fibrosis airway epithelia fail to kill bacteria because of
mobilisation and clearance of more peripherally situated abnormal airway surface fluid. Cell 1996; 85: 229–236.
secretions by RIM than that achieved by SP. This may be due 3. Matsui H, Grubb BR, Tarran R, et al. Evidence for
to the effect of sustained near maximal inspiratory efforts on pericilliary liquid layer depletion, not abnormal ion compo-
sition, in the pathogenesis of cystic fibrosis. Cell 1998; 95:
air and fluid dynamics in the distal airways. Within a
1005–1015.
distensible tube, such as the smaller airways, flow is pro- 4. Konstan MW, Hilliard KA, Norvell TM, Berger M.
portional to pressure, and the volume of fluid or air moved Bronchoalveolar lavage findings in cystic fibrosis patients
depends on the duration over which the driving pressure is with stable clinically mild lung disease suggest ongoing
exerted [28]. The RIM method requires the patient to carry infection and inflammation. Am J Respir Crit Care Med
out a full inspiratory vital capacity manoeuvre at 80% of 1994; 150: 448–454.
maximal pressure, by breathing against a fixed resistance. 5. Nixon LS, Yung B, Bell SC, Elborn JS, Shale DJ. Circulating
This leads to an extended inspiratory time secondary to the immunoreactive interleukin-6 (IL-6) in cystic fibrosis. Am
reduced airflow at the mouth. Thus, RIM may increase J Respir Crit Care Med 1998; 157: 1764–1770.
inspiratory airflow to more peripheral airways. If this occurs, 6. Konstan MW, Berger M. Current understanding of the
there may be shearing forces acting on airways secretions, inflammatory process in cystic fibrosis: onset and etiology.
causing the break up of sputum during the inspiratory phase, Pediatr Pulmonol 1997; 24: 137–142.
which might reduce the need for higher expiratory flow rates 7. Dean TP, Dai Y, Shute JK, Church MK, Warner JO.
and delay the time to reach the equal pressure point in small Interleukin-8 concentrations are elevated in bronchoalveolar
airways. The RIM method may also move the equal pressure lavage, sputum and sera of children with cystic fibrosis.
point proximally by reducing RV and enabling increased Pediatr Res 1993; 34: 159–161.
airflow with less dynamic compression from a starting point 8. Sagel SD, Kapsner R, Osberg I, Sontag MK, Accurso FJ.
of lower lung volume [29]. It may be of note that the patients Airway inflammation in children with cystic fibrosis and
healthy children assessed by sputum induction. Am J Respir
at this centre were familiar with the active cycle of breathing
Crit Care Med 2001; 15: 164: 1425–1431.
technique, the drainage and percussion techniques and the 9. Sagel SD, Sontag MK, Wagener JS, Kapsner RK, Osberg I,
RIM. Skill acquisition of both techniques is likely to be a Accurso FJ. Induced sputum inflammatory measures corre-
factor in determining the effectiveness of treatment. late with lung function in children with cystic fibrosis.
Sputum weight rather than volume was determined in this J Pediatr 2002; 141: 811–817.
study, as expectorated sputum tended to be inhomogeneous 10. Schulz C, Wolf K, Harth M, et al. Expression and release of
in nature. As physiotherapy may produce benefits through interleukin-8 by human bronchial epithelial cells from
removal of injurious products from the airways, a range of patients with chronic obstructive pulmonary disease, smokers
indicators of the pulmonary inflammatory response to and never-smokers. Respiration 2003; 70: 254–261.
chronic bacterial infection was also determined. The RIM 11. Beeh KM, Kornmann O, Buhl R, Culpitt SV, Giembycz MA,
 SPUTUM EXPECTORATION IN CF 439

Barnes PJ. Neutrophil chemotactic activity of sputum from Determinants of adherence in adults with cystic fibrosis.
patients with COPD: role of interleukin 8 and leukotriene Thorax 2002; 57: 459–464.
B4. Chest 2003; 123: 1240–1247. 21. Ionescu AA, Chatham K, Davies CA, Nixon LS, Enright S,
12. Nelson R, Miller S, Hall D, Clayton B. Chest physiotherapy Shale DJ. Inspiratory muscle function and body composition in
in cystic fibrosis. Thorax 1995; 50: 1123–1124. cystic fibrosis. Am J Respir Crit Care Med 1998; 158: 1271–1276.
13. Williams MT. Chest Physiotherapy and cystic fibrosis: why is 22. Chatham K, Baldwin J, Griffiths H, Summers L, Enright S.
the most effective form of treatment still unclear? Chest 1994; Inspiratory muscle training improves shuttle run perfor-
106: 1872–1882. mance in healthy subjects. Physiotherapy 1999; 85: 676–683.
14. Webber BA, Pryor JA. Physiotherapy techniques. In: Pryor 23. Guidelines for the measurement of respiratory function.
JA, Webber BA, eds. Physiotherapy for Respiratory and Recommendations of the British Thoracic Society and the
Cardiac Problems, 2nd Ed. Edinburgh, Churchill Living- Association of Respiratory Technicians and Physiologists.
stone, 1998; pp.137–155. Respir Med 1994; 88: 165–195.
15. Thomas J, Cook DJ, Brooks D. Chest physical therapy 24. Office of population censuses and surveys. The Dietary and
management of patients with cystic fibrosis: a meta-analysis. Nutritional Survey of British Adults 1990: 237.
Am J Respir Crit Care Med 1995; 151: 846–850. 25. Van der Schans CP. Forced expiratory manoeuvres to
16. Pryor JA, Webber BA, Hodson ME, Batten JC. Evaluation increase transport of bronchial mucus: a mechanistic
of the forced expiration technique as an adjunct to postural approach. Mondaldi Arch Chest Dis 1997; 52: 367–370.
drainage in cystic fibrosis. BMJ 1979; 2: 417–418. 26. Zach MS. The physiology of forced expiration. Paediatr
17. Lannefors L, Wollmer P. Mucus clearance with three chest Respir Rev 2000; 1: 36–39.
physiotherapy regimens in cystic fibrosis: a comparison 27. Pryor JA. Physiotherapy for airways clearance in adults. Eur
between postural drainage, PEP and physical exercise. Eur Respir J 1999; 14: 1418–1424.
Respir J 1992; 5: 748–753. 28. Regnis JA, Donnelly PM, Robinson M, Alison JA, Bye PT.
18. Apps EM, Kieselman R, Reinhardt D, et al. Sputum Ventilatory mechanics at rest and during exercise in patients
rheology changes in cystic fibrosis lung disease following with cystic fibrosis. Am J Respir Crit Care Med 1996; 154:
two different types of physiotherapy: flutter vs autogenic 1418–1425.
drainage. Chest 1998; 114: 171–177. 29. Maclem PT. The mechanics of breathing. Am J Respir Crit
19. Abbott J, Dodds M, Webb AK. Health perceptions and Care Med 1998; 157: 588–594.
treatment adherence in adults with cystic fibrosis. Thorax 30. Bonfield TL, Panuska JR, Konstan MW, et al. Inflammatory
1996; 51: 1233–1238. cytokines in cystic fibrosis lungs. Am J Respir Crit Care Med
20. Kettler LJ, SM Sawyer, HR Winefield, HW Greville. 1995; 152: 2111–2118.