- Email: [email protected]
Observations on sputum production in patients with variable airflow obstruction; Implications for the diagnosis of asthma and chronic bronchitis
Respiratory Medicine (1989) 83, 25-31
Observations on sputum production in patients with variable airflow obstruction; Implications for the diagnosis of asthma and chronic bronchitis P. J. M. OPENSHAW* AND M. TURNER-WARWICK
Cardiothoracic Institute, Brornpton Hospital, Fulham Road, London SW3 6HP, U.K.
There are few clinical studies on hypersecretion in asthma, defined as variable airflow obstruction. One hundred and thirty defined asthmatics were questioned in detail about their clinical features, with special reference to sputum production. Other clinical and laboratory data were obtained from the hospital notes. Thirty (23%) denied sputum production at any time. Of the hundred patients reporting sputum, 55 reported the largest volume when an attack ofasthma was at its worst and 42 reported most sputum during recovery. Three patients reported unchanging sputum volumes regardless of asthma attacks. There was an association between greater sputum volumes and increased absenteeism due to respiratory symptoms (P<0.01) and negative prick skin tests (P < 0.001). Cigarette smoking and bronchopulmonary aspergillosis were uncommon among our patients, and did not alone account for the sputum production rates. Sixty-eight per cent of patients reported that their pattern or volume of sputum production during attacks had not changed substantially since the onset of their asthma. Patients also fulfilling the Medical Research Council (MRC) criteria for chronic bronchitis (n = 56) were more likely to have smoked than those who did not (P < 0.005), although 32 (57%) of the asthmatics with chronic bronchitis had smoked little or not at all. Moreover the lability of peak flow measurements was similar in asthmatic patients with and without chronic bronchitis (Lability index 43 and 42% respectively). The prevalance of persistent airflow limitation (PAL) in addition to a variable component related to duration of asthma but not to smoking. The relation between PAL and sputum production was complex. The prevalence of PAL was greater in the groups of patients with slight or copious sputum and was less in the groups with intermediate sputum volumes. Our results also confirm that hypersecretion is a common feature of asthma, and support the view that asthma and chronic bronchitis are not mutually exclusive diagnoses as defined by conventional criteria. Variability of airflow obstruction should be sought and treated in patients with chronic or recurrent sputum production regardless of a history of cigarette smoking. From the bimodal relationship between sputum production and 'severe asthma' we speculate that both hypersecretion of mucus and reduced mucus clearance (perhaps associated with qualitative differences) may cause mucus to accumulate within the bronchi in chronic severe asthma, and therefore contribute to persistent airflow limitation.
Introduction Patients with episodic wheezy breathlessness together with physiological evidence of variable airflow limitation have a wide spectrum of clinical features. Some have periods of full remissioh when lung function returns to normal, whilst others, in addition to a variable component also have some persistent airflow limitation even after maximal bronchodilator therapy (1). Some patients report copious sputum production during attacks of. wheezy breathReceivedin revisedform March 1988. * Present address: National Institute for Medical Research, The Ridgeway,Mill Hill, London NW7 IAA, U.K. 0954-61 ! I/89/010025 + 08 $03.00/0
lessness, whilst others complain mainly of wheeze and dyspnoea. In this report the word asthma is used to describe patients with episodic wheezy breathlessness and variable airflow limitation without prejudice to the implications of conventional diagnostic disease entities. Since asthma has been defined in terms of variable airflow limitation (3), much attention has been focussed on measurement of airway calibre. Sputum production in asthmatics has received relatively little attention, despite the finding that patients with a history of chronic asthma in apparent remission have widespread mucus plugging of intrapulmonary airways (4-6). It has been postulated that accumulation © 1989 Bailli&e Tindall
26
P. J. M. Openshaw and M. Turner- Warwick
of intrabronchial secretions may contribute to the 'fixed' element of airflow obstruction, often present in persisting asthma (1,7), especially in those without radiographic or physiological evidence of destructive emphysema. Chronic bronchitis has been defined by the Medical Research Council in terms of cough and sputum production originally for epidemiological purposes (8). Although there is increasing awareness of the overlap between the diagnosis of asthma and chronic bronchitis in individual patients, many physicians still frequently classify patients as having either chronic bronchitis (i.e. symptoms of cough and sputum) or asthma (i.e. symptoms of episodic wheeziness and physiological measurements showing variation in airflow). To assume that these categories are mutually exclusive may lead to undertreatment of asthma. The aim of this study was to establish the prevalence of acute and chronic sputum production in a group of carefully defined patients with asthma. In addition, we sought relationships between the pattern of sputum production and other clinical features, including the presence of an additional 'fixed' component of airflow limitation. Our results highlight some areas of overlap between groups of patients who might conventionally be allocated to different diagnostic categories, and show that asthmatics demonstrate heterogeneity concerning sputum production. Methods SELECTION OF PATIENTS
One hundred and forty-five patients aged between ten and 60 years attending the Brompton or New Cross asthma clinics and giving a clear history of episodic wheezy breathlessness were specially interviewed during June and July 1983 and invited to complete a detailed questionnaire. In all other respects they were unselected. In particular, some were new patients and others interviewed during tI']eir follow-up attendances. Of these patients, eight were later excluded because of radiographic features of bronchiectasis or pulmonary fibrosis. Seven were excluded from analysis because they failed to complete the questionnaire, or because there was missing data in the hospital records. Thus, 130 patients were accepted into the study. A bronchial lability index was calculated for each patient by expressing the greatest documented change (either spontaneously or following treatment) in peak expiratory flow rate (PEFR) as a percentage of the patient's normal P E F R (i.e. Lability I n d e x = m a x i mum change in PEFR/predicted normal P E F R × 100) as derived from the nomograms of Gregg and Nunn
(10). One hundred and nine out of 130 patients had a lability index of greater than 15%. Of the 21 with a lesser lability index all gave a clear history of repeated episodes of wheezy breathlessness with a reported immediate response to beta2 adrenoceptor agonists and, therefore, fulfilled the eligibility criteria but were mainly new attenders in whom serial peak flow measurements were not obtained, and at the time of interview, had spirometric measurements within the normal range. We did not exclude these patients from the study because it would have introduced a potential bias towards more severe asthmatics who had attended the clinics more frequently. Some clinical characteristics of the 130 patients studied are shown in Table 1. THE QUESTIONNAIRE The questionnaire consisted of 92 questions relating to the patient's asthma. Of these,-28 were about sputum production. Questions which could be answered subjectively were asked in different ways in different sections of the questionnaire, so that inconsistent answers could be detected when the interviewer went through the questionnaire with the patient. All patients were interviewed by the same clinician, and took 45 to 120 rain to answer all the questions. The results of clinical and laboratory tests were obtained from each patient's notes. Sputum production The questions used to grade the pattern and quantity of sputum production were: During the past year, how much sputum do you cough up each day: !. When you are well (at your best)? 2. When a typical asthma attack is at its worst? 3. When you are starting to recover from an attack? In answer to each question patients were asked to Table 1 Some clinical characteristics of the 130 asthmatics, completing the questionnaire
Females Smokers* Positive skin testst Hay fever Bronchopulmonary aspergillosis~ Mean age at interview (yr+so) Mean age at onset of asthma (yr +SD)
83 (64°/,,) 38 (29°/,,) 97 (75°/,,) 56 (43%) 14 (11%) 35'4+ 13"6 17'2+ 15-4
* More than 1 pack year. t Immediate prick skin tests to more than I allergen from a standard panel of 21 common environmental agents. BPA defined according to criteria of Malo et al. (9).
Observations on sputum production choose their usual sputum production 'rate' (i.e. volume in a 24 h period) from the five following categories: None (0 ml) Few 'blobs" (approx. 2 ml) Teaspoonful (approx. 5 ml) Eggcupful (approx. 20 ml) Teacup or more (approx. 150 ml) Direct measurements of sputum volumes were not undertaken for several reasons. At best such measurements could only be attempted over short periods of time while this study aims to assess sputum production over the course of the natural history of the disorder. Sputum collections in a working population was impractical and the completeness of the collections would remain unknown. A previous attempt to collect sputum to validate questionnaire data had failed (1 I).
Chronic bronchitis Patients reporting sputum (but without reference to volume) on most days for a total of 3 months or more each year for at least 2 years were classed as having associated chronic bronchitis in accordance with the M R C definition. Patients reporting regular production of small volumes of sputum which they swallowed, were included in this definition. Persistent airflow limitation (PAL) Peak flow readings were taken from the hospital notes, and the number of readings necessarily varied between patients. Three or more readings over a period o f a t least a year were recorded for 120 patients, although most of these had many more available readings. Patients who, in addition to demonstrating variability in airflow limitation showed all recorded P E F R readings, even after bronchodilators, of less than 80% of the predicted value were defined as having persistent airflow limitation. Severe asthma • Patients with asthma severe enough to cause absenteeism from work or school, or periods when normal daily tasks could not be performed, for two or more weeks in the past year and PAL were classed as 'severe asthmatics'. Cigarette smoking Patients who had smoked more than one pack year were classed as 'smokers' regardless of current smoking habit. Thirty eight of 130 (29%) were thus classed. Skin tests
Skin prick tests to a standard range o f 21 common
27
environmental antigens were used. A wheal size of greater than 2 mm read at I 0-15 min was regarded as a positive response.
Statistics A Perkin-Elmer 3220 computer was used for statistical analysis with the Minitab (Pennsylvania State University) system. The Kruskal- Wallis test (equivalent to Mann-Whitney-U analysis of the rank order of a variable) was used when data did not have a normal distribution. Results The mean age of the population at the time of study was 35-4 4- 13.6 yr SD and the age of onset of wheezy attacks was 17.2+ 15.4 yr SD. Ninety eight of 130 (75%) patients showed positive skin reactions (Table 1). One hundred of 130 (77%) patients reported sputum production. THE PATTERN OF SPUTUM PRODUCTION SINCE THE ONSET OF ASTHMA
Eighty-eight of 130 (68%) patients said that there had been no substantial change in the presence or absence of sputum production, or its approximate volume per day, since the onset of their asthmatic symptoms. Three of the 30 patients who currently produced no sputum claimed to have originally produced sputum with attacks. Of those reporting sputum 28 of 100 said that the sputum production during attacks had increased over the years. Most patients who said that production had increased, attributed the change to a spontaneous deterioration of their asthma. SPUTUM PRODUCTION IN RELATION TO ASTHMA ATTACKS
Thirty (23%) patients produced no sputum at any time during attacks of asthma. Of the 100 patients reporting sputum, 55 reported most when the attack was at its worst and 42 most during recovery. Three reported similar amounts of sputum production regardless of asthmatic attacks. MORBIDITY AND SPUTUM PRODUCTION RATE There was a statistically significant increase in time off work or school in patients with copious sputum production (P < 0.01, Kruskal-Wallis test; see Fig. 1). A similar trend was evident in the frequency of asthma-related calls to the general practitioner, frequency of hospital admissions and duration of asthma attacks, although these trends were not statistically significant, all of these associations were unaffected by removal o f patients with BPA from the analysis.
28
P. J. M. Openshaw and M. Turner- Warwick 90 80
70 60 m 50 40 '- 2-.~_
30 20
0
~
n =30 0
27
20
32
21
2
5
20
~>150
10
Approximate peak daily sputum volume (ml)
0
Fig. 1 Time offwork (or school, ortimeofincapacityifnot working) during the past year according to maximum sputum production rates.
n=37 0-9
49 10-19
19
13
12
20-29:30-39
>39
Duration of asthma (years)
Fig. 2 Persistent airflow obstruction in relation to duration of asthma.
Sputum volumes were also associated with negative skin test responses (P<0.001) but not to a negative family history of atopy. Larger sputum volumes were not significantly related to cigarette smoking. PERSISTENTAIRFLOWLIMITATION(PAL) In 52 (40%) of our 130 patients, all the recorded peak flows were less than 80% of the predicted value. The change in prevalance of persistent airflow limitation (PAL) with duration of asthma is shown in Fig. 2. Although older patients tend to have a longer history of asthma, there was no relationship between age per s~ and PAL. PAL was also unrelated to cigarette smoking, which was largely confined to those with less frequent asthma attacks or asthmatics of recent onset. PERSISTENT AIRFLOW LIMITATION AND SPUTUM RATE
PAL was commonest in groups of patients with the least amount as well as the most sputum (Fig. 3); Statistically this association was not significant. Patients with both PAL and absenteeism for two or more weeks in the past year (i.e. 'severe asthmatics') were also more common in the groups representing the extremes of sputum production (P < 0-01) (Fig. 4). ASSOCIATED CHRONIC BRONCHITIS
Of 130 patients, 56 (43%) had sufficiently prolonged or recurrent sputum production (irrespective of changes in volume during attacks of asthma) to fulfil the criteria for chronic bronchitis (see Methods).
60
50
40 _J
.< It-
30
20
10
0
n =30 0
27 2
20 5
32 20
21 ~>150
Approximate peak daily sputum volume (ml)
Fig. 3 The percentage ofpatients in each sputum group with partial airflow limitation (PAL).
Observations on sputum production 50
._1
<
0,,. 13 tI'o
E ¢-
30
(Io e) >
20
r"
10
n =30 0
27
20
32
21
2 5 20 1>150 Approximate peak daily sputum v o l u m e (ml)
Fig. 4 The percentage of patients in each sputum group with severe asthma and persistent airflow obstruction (defined in text).
Notably however, the lability index of PEFR was similar in patients with and without chronic bronchitis (43 +__26% and 42___23% respectively) and PAL was equally frequent. Patients with chronic bronchitis were more likely to have been smokers (24 our of 56 versus 14 out of 74; P < 0.005) but sputum volume was unrelated to cigarette smoking. On the other hand, 32 (57%) of patients with associated chronic bronchitis were non-smokers. Discussion
Hypersecretion in asthma is important not only as a potential contributing factor to reduction in airway calibre, but inspissated sputum is an invariable feature of fatal asthma (4,5). There is much experimental work suggesting factors which promote hypersecretion in asthma and the therapeutic agents which may r~luce it. However, there are surprisingly few studies attempting to define the clinical characteristics of sputum production either in relation to attacks or over the time course of the natural history of asthma in individual patients. The aim o f this study was to establish the prevalence and pattern of sputum production in a group of patients with clearly defined asthma attending hospital
29
clinics, and to see whether the rate of sputum production (i.e. volume in a 24 h period) was associated with increased morbidity, persistent airflow limitation, smoking or other clinical factors. Any clinical study of sputum production is difficult because there is no absolute way of validating the method used to quantify sputum production. Sputum collections over prolonged periods of time have been shown to be incomplete in a working outpatient population (I 1) and for involuntary or social reasons some sputum is usually swallowed. Patient interview has been shown in other studies to be as good (or bad) as any other method in studies on chronic bronchitis (1 I). For these reasons sputum collections were not attempted. The design of the questionnaire was however able to test the reproducibility of replies and showed that these were very consistent, both when the same question was presented in different ways within the questionnaire and when the questionnaire was completed on a separate occasion in eight patients. Most patients had surprisingly little difficulty in recalling sputum production and assigning themselves to one of the categories regarding sputum volume indicated in the questionnaire. Our results showed that many patients confirmed the generally held view that expectoration is difficult at the height of an asthma attack but increases as the attack wains. However, contrary to popular belief 55% of those producing sputum reported maximum volumes at the height of an attack. These results, if confirmed, might suggest that there are certain trigger factors or mechanisms of airway narrowing which are associated with hypersecretion and others which are not. It is often supposed that as the duration of asthma increases so variability of airway calibre lessens and regular sputum production increases. However, the study suggests that in the majority of patients there is little change in the amount of sputum produced over time. Some produce sputum from the beginning of their asthmatic symptoms and others produce no sputum even after many years. Clearly these reports depend entirely on patient recall. However, we were impressed by how easily patients allocated themselves to different categories regarding sputum production over long periods of time. It is also of course possible that medication modifies sputum production in individual patients. This indeed was shown to be the case in one long term study on inhaled beclomethasone diproprionate (12). Nevertheless, the range of antiasthma medication being taken at the time of the presem study was no different in the group o f nonsecretory patients compared to those with substantial secretions.
30
P. J. M. Openshaw and M. Turner- Warwick
Although there is increasing awareness to the contrary, there are still many physicians who assume that asthma, defined in terms of functional measurements of airways resistance, and chronic bronchitis, defined in terms of symptoms of cough and sputum, are mutually exclusive. There is no logical reason why this should be so when the diagnostic criteria are based on different modalities. The present study supports the view that not only do the majority of patients fulfilling standards criteria for asthma also produce sputum (100 out of 130) but that 54% of those producing sputum do so sufficiently consistently also to fulfill the conventional criteria for chronic bronchitis as set out by the Medical Research Council (8). It is important to emphasise that 75% of patients included in our study of patients with episodic wheezy breathlessness were atopic in terms of a positive prick skin test and the mean age of onset of symptoms was 17 years. These facts provide further supporting evidence that our patients had other features typical of an asthmatic population (13). In the present study not only was bronchial lability substantially increased above normal in those with and without associated chronic bronchitis but there was no difference in the degree of increased lability in the two groups. In view of the similar lability indices it seems improbable that histamine or methacholine challenge tests for bronchial hyper reactivity would discriminate further, but this was not formally tested. This study provides further evidence to support the increasingly recognized view that the categorising of patients into those with variable airflow obstruction or sputum production is not justified and is misleading because it may lead to undertreatment of patients with a reversible component of airflow obstruction. A component of persistent airflow limitation in patients who also have continuing substantial variability in airway calibre is now well recognized (7,14) even in those on maximum treatment (I). The p athophysioiogy of this is ill understood. Our study shows that PAL is not limited to those with increased sputum volumes but is more common in the groups characterized by the extremes of sputum production (i.e. either none or much sputum). PAL was found much less frequently in those with small sputum volumes. This observation had led us to the hypothesis that clearance rates and/or sputum quality may be important contributing factors to PAL. Where clearance is impaired expectoration will diminish and the increased retention o f secretions with associated bronchial epithelial damage could increase PAL. Where excessive secretion rates overwhelm a normal clearance, this too could lead to sputum retention and PAL. On the other hand where secretion and clearance rates are in
balance, small sputum volumes without PAL might be expected. This hypothesis could explain the bimodial distribution curve seen in Figs 3 and 4. The heterogeneity of the asthma response is now well recognized. Examples include atopic as well as non atopic subjects, those with and without associated blood eosinophilia or raised lgE levels and childhood versus late onset types. The heterogeneity of mechanisms of asthma is also increasingly recognised. The observations made in our study suggest that yet another type of heterogeneity may be characterized by the q u a n t i t y and perhaps the quality of sputum produced. The mechanism of sputum production and the explanation for its variations should now be explored more fully.
Acknowledgement We thank Professor Charles Fletcher, Dr Janet Derbyshire and Mr Andrew Nunn for advice about the design of the questionnaire, Professors Peter Cole, T. J. H. Clark, A. B. Kay and Dr G. M. Cochrane for allowing us to include their patients in the survey and Shering Pharmaceuticals for financial support.
References 1. Brown PJ, Greville HW, Finucane KE. Asthma and irreversible airflow obstruction. Thoeax 1984; 39: 131136. 2. Kay AB. The sputum in bronchial asthma. In: Clark TJH, Godfrey S, eds: Asthma, 2nd Edition. London: Chapman and Hall, 1983. 3. Ciba Foundation Study Group No. 38. Porter R, Birch J eds. Identification of Asthma. Edinburgh: Churchill Livingstone, 1971. 4. Dunnill MS. The pathology of asthma. In: Transactions of the WorM Asthma Conference. London: The Chest and Heart Association, 1975. 5. Glynn AA, Michaels L. Bronchial biopsy in chronic bronchitis and asthma. Thorax 1960; 15: 142-51. 6. Thurlbeck WM. In: Chronic Airflow Obstruction in Lung Disease. Philadelphia: WB Saunders, 1976. 7. Tschopp JM, Turner-Warwick M. Persistent airflow obstruction in asthmatics receiving routine self-adjusted medication. Eur J Respir Dis 1984; 65: 346-353. 8. Medical Research Council. Definition and classification of chronic bronchitis for clinical and epidemiological purposes. Lancet 1965; i: 776-779. 9. Malo JL, Hawkins, Pepys J. Studies in chronic allergic bronchopulmonary aspergillosis. Thorax 1977; 32: 254261. 10. Gregg I, Nunn AJ. Peak expiratory flow in normal subjects. BR Med J 1973; 3: 282-284. I I. Fletcher C, Peto R, Tinker CM. A comparison of the assessment of simple bronchitis (chronic mucus hypersecretion) by measurements of sputum volume and by standard questions on phlegm production, lnt J Epidemiol 1974; 3: 315-319.
Observations on sputum production
12. Brompton Hospital/Medical Research Council Collaborative trial: double blind trial comparing two dosage schedules of beclomethasone diproprionate aerosol in chronic bronchial asthma. Chest 1979; 73: 121. 13. Hendrick DJ, Davies RJ, D'Souza MF, Pepys J. An
31
analysis of skin prick tests in 656 asthmatic patients. Thorax 1975, 30: 2. 14. Peat JK, Woolcock AJ, Cullen K. Rate ofdecline of lung function in subjects with asthma. Eur J Respir Dis 1987, 70: 171-179.
Observations on sputum production in patients with variable airflow obstruction; Implications for the diagnosis of asthma and chronic bronchitis P. J. M. OPENSHAW* AND M. TURNER-WARWICK
Cardiothoracic Institute, Brornpton Hospital, Fulham Road, London SW3 6HP, U.K.
There are few clinical studies on hypersecretion in asthma, defined as variable airflow obstruction. One hundred and thirty defined asthmatics were questioned in detail about their clinical features, with special reference to sputum production. Other clinical and laboratory data were obtained from the hospital notes. Thirty (23%) denied sputum production at any time. Of the hundred patients reporting sputum, 55 reported the largest volume when an attack ofasthma was at its worst and 42 reported most sputum during recovery. Three patients reported unchanging sputum volumes regardless of asthma attacks. There was an association between greater sputum volumes and increased absenteeism due to respiratory symptoms (P<0.01) and negative prick skin tests (P < 0.001). Cigarette smoking and bronchopulmonary aspergillosis were uncommon among our patients, and did not alone account for the sputum production rates. Sixty-eight per cent of patients reported that their pattern or volume of sputum production during attacks had not changed substantially since the onset of their asthma. Patients also fulfilling the Medical Research Council (MRC) criteria for chronic bronchitis (n = 56) were more likely to have smoked than those who did not (P < 0.005), although 32 (57%) of the asthmatics with chronic bronchitis had smoked little or not at all. Moreover the lability of peak flow measurements was similar in asthmatic patients with and without chronic bronchitis (Lability index 43 and 42% respectively). The prevalance of persistent airflow limitation (PAL) in addition to a variable component related to duration of asthma but not to smoking. The relation between PAL and sputum production was complex. The prevalence of PAL was greater in the groups of patients with slight or copious sputum and was less in the groups with intermediate sputum volumes. Our results also confirm that hypersecretion is a common feature of asthma, and support the view that asthma and chronic bronchitis are not mutually exclusive diagnoses as defined by conventional criteria. Variability of airflow obstruction should be sought and treated in patients with chronic or recurrent sputum production regardless of a history of cigarette smoking. From the bimodal relationship between sputum production and 'severe asthma' we speculate that both hypersecretion of mucus and reduced mucus clearance (perhaps associated with qualitative differences) may cause mucus to accumulate within the bronchi in chronic severe asthma, and therefore contribute to persistent airflow limitation.
Introduction Patients with episodic wheezy breathlessness together with physiological evidence of variable airflow limitation have a wide spectrum of clinical features. Some have periods of full remissioh when lung function returns to normal, whilst others, in addition to a variable component also have some persistent airflow limitation even after maximal bronchodilator therapy (1). Some patients report copious sputum production during attacks of. wheezy breathReceivedin revisedform March 1988. * Present address: National Institute for Medical Research, The Ridgeway,Mill Hill, London NW7 IAA, U.K. 0954-61 ! I/89/010025 + 08 $03.00/0
lessness, whilst others complain mainly of wheeze and dyspnoea. In this report the word asthma is used to describe patients with episodic wheezy breathlessness and variable airflow limitation without prejudice to the implications of conventional diagnostic disease entities. Since asthma has been defined in terms of variable airflow limitation (3), much attention has been focussed on measurement of airway calibre. Sputum production in asthmatics has received relatively little attention, despite the finding that patients with a history of chronic asthma in apparent remission have widespread mucus plugging of intrapulmonary airways (4-6). It has been postulated that accumulation © 1989 Bailli&e Tindall
26
P. J. M. Openshaw and M. Turner- Warwick
of intrabronchial secretions may contribute to the 'fixed' element of airflow obstruction, often present in persisting asthma (1,7), especially in those without radiographic or physiological evidence of destructive emphysema. Chronic bronchitis has been defined by the Medical Research Council in terms of cough and sputum production originally for epidemiological purposes (8). Although there is increasing awareness of the overlap between the diagnosis of asthma and chronic bronchitis in individual patients, many physicians still frequently classify patients as having either chronic bronchitis (i.e. symptoms of cough and sputum) or asthma (i.e. symptoms of episodic wheeziness and physiological measurements showing variation in airflow). To assume that these categories are mutually exclusive may lead to undertreatment of asthma. The aim of this study was to establish the prevalence of acute and chronic sputum production in a group of carefully defined patients with asthma. In addition, we sought relationships between the pattern of sputum production and other clinical features, including the presence of an additional 'fixed' component of airflow limitation. Our results highlight some areas of overlap between groups of patients who might conventionally be allocated to different diagnostic categories, and show that asthmatics demonstrate heterogeneity concerning sputum production. Methods SELECTION OF PATIENTS
One hundred and forty-five patients aged between ten and 60 years attending the Brompton or New Cross asthma clinics and giving a clear history of episodic wheezy breathlessness were specially interviewed during June and July 1983 and invited to complete a detailed questionnaire. In all other respects they were unselected. In particular, some were new patients and others interviewed during tI']eir follow-up attendances. Of these patients, eight were later excluded because of radiographic features of bronchiectasis or pulmonary fibrosis. Seven were excluded from analysis because they failed to complete the questionnaire, or because there was missing data in the hospital records. Thus, 130 patients were accepted into the study. A bronchial lability index was calculated for each patient by expressing the greatest documented change (either spontaneously or following treatment) in peak expiratory flow rate (PEFR) as a percentage of the patient's normal P E F R (i.e. Lability I n d e x = m a x i mum change in PEFR/predicted normal P E F R × 100) as derived from the nomograms of Gregg and Nunn
(10). One hundred and nine out of 130 patients had a lability index of greater than 15%. Of the 21 with a lesser lability index all gave a clear history of repeated episodes of wheezy breathlessness with a reported immediate response to beta2 adrenoceptor agonists and, therefore, fulfilled the eligibility criteria but were mainly new attenders in whom serial peak flow measurements were not obtained, and at the time of interview, had spirometric measurements within the normal range. We did not exclude these patients from the study because it would have introduced a potential bias towards more severe asthmatics who had attended the clinics more frequently. Some clinical characteristics of the 130 patients studied are shown in Table 1. THE QUESTIONNAIRE The questionnaire consisted of 92 questions relating to the patient's asthma. Of these,-28 were about sputum production. Questions which could be answered subjectively were asked in different ways in different sections of the questionnaire, so that inconsistent answers could be detected when the interviewer went through the questionnaire with the patient. All patients were interviewed by the same clinician, and took 45 to 120 rain to answer all the questions. The results of clinical and laboratory tests were obtained from each patient's notes. Sputum production The questions used to grade the pattern and quantity of sputum production were: During the past year, how much sputum do you cough up each day: !. When you are well (at your best)? 2. When a typical asthma attack is at its worst? 3. When you are starting to recover from an attack? In answer to each question patients were asked to Table 1 Some clinical characteristics of the 130 asthmatics, completing the questionnaire
Females Smokers* Positive skin testst Hay fever Bronchopulmonary aspergillosis~ Mean age at interview (yr+so) Mean age at onset of asthma (yr +SD)
83 (64°/,,) 38 (29°/,,) 97 (75°/,,) 56 (43%) 14 (11%) 35'4+ 13"6 17'2+ 15-4
* More than 1 pack year. t Immediate prick skin tests to more than I allergen from a standard panel of 21 common environmental agents. BPA defined according to criteria of Malo et al. (9).
Observations on sputum production choose their usual sputum production 'rate' (i.e. volume in a 24 h period) from the five following categories: None (0 ml) Few 'blobs" (approx. 2 ml) Teaspoonful (approx. 5 ml) Eggcupful (approx. 20 ml) Teacup or more (approx. 150 ml) Direct measurements of sputum volumes were not undertaken for several reasons. At best such measurements could only be attempted over short periods of time while this study aims to assess sputum production over the course of the natural history of the disorder. Sputum collections in a working population was impractical and the completeness of the collections would remain unknown. A previous attempt to collect sputum to validate questionnaire data had failed (1 I).
Chronic bronchitis Patients reporting sputum (but without reference to volume) on most days for a total of 3 months or more each year for at least 2 years were classed as having associated chronic bronchitis in accordance with the M R C definition. Patients reporting regular production of small volumes of sputum which they swallowed, were included in this definition. Persistent airflow limitation (PAL) Peak flow readings were taken from the hospital notes, and the number of readings necessarily varied between patients. Three or more readings over a period o f a t least a year were recorded for 120 patients, although most of these had many more available readings. Patients who, in addition to demonstrating variability in airflow limitation showed all recorded P E F R readings, even after bronchodilators, of less than 80% of the predicted value were defined as having persistent airflow limitation. Severe asthma • Patients with asthma severe enough to cause absenteeism from work or school, or periods when normal daily tasks could not be performed, for two or more weeks in the past year and PAL were classed as 'severe asthmatics'. Cigarette smoking Patients who had smoked more than one pack year were classed as 'smokers' regardless of current smoking habit. Thirty eight of 130 (29%) were thus classed. Skin tests
Skin prick tests to a standard range o f 21 common
27
environmental antigens were used. A wheal size of greater than 2 mm read at I 0-15 min was regarded as a positive response.
Statistics A Perkin-Elmer 3220 computer was used for statistical analysis with the Minitab (Pennsylvania State University) system. The Kruskal- Wallis test (equivalent to Mann-Whitney-U analysis of the rank order of a variable) was used when data did not have a normal distribution. Results The mean age of the population at the time of study was 35-4 4- 13.6 yr SD and the age of onset of wheezy attacks was 17.2+ 15.4 yr SD. Ninety eight of 130 (75%) patients showed positive skin reactions (Table 1). One hundred of 130 (77%) patients reported sputum production. THE PATTERN OF SPUTUM PRODUCTION SINCE THE ONSET OF ASTHMA
Eighty-eight of 130 (68%) patients said that there had been no substantial change in the presence or absence of sputum production, or its approximate volume per day, since the onset of their asthmatic symptoms. Three of the 30 patients who currently produced no sputum claimed to have originally produced sputum with attacks. Of those reporting sputum 28 of 100 said that the sputum production during attacks had increased over the years. Most patients who said that production had increased, attributed the change to a spontaneous deterioration of their asthma. SPUTUM PRODUCTION IN RELATION TO ASTHMA ATTACKS
Thirty (23%) patients produced no sputum at any time during attacks of asthma. Of the 100 patients reporting sputum, 55 reported most when the attack was at its worst and 42 most during recovery. Three reported similar amounts of sputum production regardless of asthmatic attacks. MORBIDITY AND SPUTUM PRODUCTION RATE There was a statistically significant increase in time off work or school in patients with copious sputum production (P < 0.01, Kruskal-Wallis test; see Fig. 1). A similar trend was evident in the frequency of asthma-related calls to the general practitioner, frequency of hospital admissions and duration of asthma attacks, although these trends were not statistically significant, all of these associations were unaffected by removal o f patients with BPA from the analysis.
28
P. J. M. Openshaw and M. Turner- Warwick 90 80
70 60 m 50 40 '- 2-.~_
30 20
0
~
n =30 0
27
20
32
21
2
5
20
~>150
10
Approximate peak daily sputum volume (ml)
0
Fig. 1 Time offwork (or school, ortimeofincapacityifnot working) during the past year according to maximum sputum production rates.
n=37 0-9
49 10-19
19
13
12
20-29:30-39
>39
Duration of asthma (years)
Fig. 2 Persistent airflow obstruction in relation to duration of asthma.
Sputum volumes were also associated with negative skin test responses (P<0.001) but not to a negative family history of atopy. Larger sputum volumes were not significantly related to cigarette smoking. PERSISTENTAIRFLOWLIMITATION(PAL) In 52 (40%) of our 130 patients, all the recorded peak flows were less than 80% of the predicted value. The change in prevalance of persistent airflow limitation (PAL) with duration of asthma is shown in Fig. 2. Although older patients tend to have a longer history of asthma, there was no relationship between age per s~ and PAL. PAL was also unrelated to cigarette smoking, which was largely confined to those with less frequent asthma attacks or asthmatics of recent onset. PERSISTENT AIRFLOW LIMITATION AND SPUTUM RATE
PAL was commonest in groups of patients with the least amount as well as the most sputum (Fig. 3); Statistically this association was not significant. Patients with both PAL and absenteeism for two or more weeks in the past year (i.e. 'severe asthmatics') were also more common in the groups representing the extremes of sputum production (P < 0-01) (Fig. 4). ASSOCIATED CHRONIC BRONCHITIS
Of 130 patients, 56 (43%) had sufficiently prolonged or recurrent sputum production (irrespective of changes in volume during attacks of asthma) to fulfil the criteria for chronic bronchitis (see Methods).
60
50
40 _J
.< It-
30
20
10
0
n =30 0
27 2
20 5
32 20
21 ~>150
Approximate peak daily sputum volume (ml)
Fig. 3 The percentage ofpatients in each sputum group with partial airflow limitation (PAL).
Observations on sputum production 50
._1
<
0,,. 13 tI'o
E ¢-
30
(Io e) >
20
r"
10
n =30 0
27
20
32
21
2 5 20 1>150 Approximate peak daily sputum v o l u m e (ml)
Fig. 4 The percentage of patients in each sputum group with severe asthma and persistent airflow obstruction (defined in text).
Notably however, the lability index of PEFR was similar in patients with and without chronic bronchitis (43 +__26% and 42___23% respectively) and PAL was equally frequent. Patients with chronic bronchitis were more likely to have been smokers (24 our of 56 versus 14 out of 74; P < 0.005) but sputum volume was unrelated to cigarette smoking. On the other hand, 32 (57%) of patients with associated chronic bronchitis were non-smokers. Discussion
Hypersecretion in asthma is important not only as a potential contributing factor to reduction in airway calibre, but inspissated sputum is an invariable feature of fatal asthma (4,5). There is much experimental work suggesting factors which promote hypersecretion in asthma and the therapeutic agents which may r~luce it. However, there are surprisingly few studies attempting to define the clinical characteristics of sputum production either in relation to attacks or over the time course of the natural history of asthma in individual patients. The aim o f this study was to establish the prevalence and pattern of sputum production in a group of patients with clearly defined asthma attending hospital
29
clinics, and to see whether the rate of sputum production (i.e. volume in a 24 h period) was associated with increased morbidity, persistent airflow limitation, smoking or other clinical factors. Any clinical study of sputum production is difficult because there is no absolute way of validating the method used to quantify sputum production. Sputum collections over prolonged periods of time have been shown to be incomplete in a working outpatient population (I 1) and for involuntary or social reasons some sputum is usually swallowed. Patient interview has been shown in other studies to be as good (or bad) as any other method in studies on chronic bronchitis (1 I). For these reasons sputum collections were not attempted. The design of the questionnaire was however able to test the reproducibility of replies and showed that these were very consistent, both when the same question was presented in different ways within the questionnaire and when the questionnaire was completed on a separate occasion in eight patients. Most patients had surprisingly little difficulty in recalling sputum production and assigning themselves to one of the categories regarding sputum volume indicated in the questionnaire. Our results showed that many patients confirmed the generally held view that expectoration is difficult at the height of an asthma attack but increases as the attack wains. However, contrary to popular belief 55% of those producing sputum reported maximum volumes at the height of an attack. These results, if confirmed, might suggest that there are certain trigger factors or mechanisms of airway narrowing which are associated with hypersecretion and others which are not. It is often supposed that as the duration of asthma increases so variability of airway calibre lessens and regular sputum production increases. However, the study suggests that in the majority of patients there is little change in the amount of sputum produced over time. Some produce sputum from the beginning of their asthmatic symptoms and others produce no sputum even after many years. Clearly these reports depend entirely on patient recall. However, we were impressed by how easily patients allocated themselves to different categories regarding sputum production over long periods of time. It is also of course possible that medication modifies sputum production in individual patients. This indeed was shown to be the case in one long term study on inhaled beclomethasone diproprionate (12). Nevertheless, the range of antiasthma medication being taken at the time of the presem study was no different in the group o f nonsecretory patients compared to those with substantial secretions.
30
P. J. M. Openshaw and M. Turner- Warwick
Although there is increasing awareness to the contrary, there are still many physicians who assume that asthma, defined in terms of functional measurements of airways resistance, and chronic bronchitis, defined in terms of symptoms of cough and sputum, are mutually exclusive. There is no logical reason why this should be so when the diagnostic criteria are based on different modalities. The present study supports the view that not only do the majority of patients fulfilling standards criteria for asthma also produce sputum (100 out of 130) but that 54% of those producing sputum do so sufficiently consistently also to fulfill the conventional criteria for chronic bronchitis as set out by the Medical Research Council (8). It is important to emphasise that 75% of patients included in our study of patients with episodic wheezy breathlessness were atopic in terms of a positive prick skin test and the mean age of onset of symptoms was 17 years. These facts provide further supporting evidence that our patients had other features typical of an asthmatic population (13). In the present study not only was bronchial lability substantially increased above normal in those with and without associated chronic bronchitis but there was no difference in the degree of increased lability in the two groups. In view of the similar lability indices it seems improbable that histamine or methacholine challenge tests for bronchial hyper reactivity would discriminate further, but this was not formally tested. This study provides further evidence to support the increasingly recognized view that the categorising of patients into those with variable airflow obstruction or sputum production is not justified and is misleading because it may lead to undertreatment of patients with a reversible component of airflow obstruction. A component of persistent airflow limitation in patients who also have continuing substantial variability in airway calibre is now well recognized (7,14) even in those on maximum treatment (I). The p athophysioiogy of this is ill understood. Our study shows that PAL is not limited to those with increased sputum volumes but is more common in the groups characterized by the extremes of sputum production (i.e. either none or much sputum). PAL was found much less frequently in those with small sputum volumes. This observation had led us to the hypothesis that clearance rates and/or sputum quality may be important contributing factors to PAL. Where clearance is impaired expectoration will diminish and the increased retention o f secretions with associated bronchial epithelial damage could increase PAL. Where excessive secretion rates overwhelm a normal clearance, this too could lead to sputum retention and PAL. On the other hand where secretion and clearance rates are in
balance, small sputum volumes without PAL might be expected. This hypothesis could explain the bimodial distribution curve seen in Figs 3 and 4. The heterogeneity of the asthma response is now well recognized. Examples include atopic as well as non atopic subjects, those with and without associated blood eosinophilia or raised lgE levels and childhood versus late onset types. The heterogeneity of mechanisms of asthma is also increasingly recognised. The observations made in our study suggest that yet another type of heterogeneity may be characterized by the q u a n t i t y and perhaps the quality of sputum produced. The mechanism of sputum production and the explanation for its variations should now be explored more fully.
Acknowledgement We thank Professor Charles Fletcher, Dr Janet Derbyshire and Mr Andrew Nunn for advice about the design of the questionnaire, Professors Peter Cole, T. J. H. Clark, A. B. Kay and Dr G. M. Cochrane for allowing us to include their patients in the survey and Shering Pharmaceuticals for financial support.
References 1. Brown PJ, Greville HW, Finucane KE. Asthma and irreversible airflow obstruction. Thoeax 1984; 39: 131136. 2. Kay AB. The sputum in bronchial asthma. In: Clark TJH, Godfrey S, eds: Asthma, 2nd Edition. London: Chapman and Hall, 1983. 3. Ciba Foundation Study Group No. 38. Porter R, Birch J eds. Identification of Asthma. Edinburgh: Churchill Livingstone, 1971. 4. Dunnill MS. The pathology of asthma. In: Transactions of the WorM Asthma Conference. London: The Chest and Heart Association, 1975. 5. Glynn AA, Michaels L. Bronchial biopsy in chronic bronchitis and asthma. Thorax 1960; 15: 142-51. 6. Thurlbeck WM. In: Chronic Airflow Obstruction in Lung Disease. Philadelphia: WB Saunders, 1976. 7. Tschopp JM, Turner-Warwick M. Persistent airflow obstruction in asthmatics receiving routine self-adjusted medication. Eur J Respir Dis 1984; 65: 346-353. 8. Medical Research Council. Definition and classification of chronic bronchitis for clinical and epidemiological purposes. Lancet 1965; i: 776-779. 9. Malo JL, Hawkins, Pepys J. Studies in chronic allergic bronchopulmonary aspergillosis. Thorax 1977; 32: 254261. 10. Gregg I, Nunn AJ. Peak expiratory flow in normal subjects. BR Med J 1973; 3: 282-284. I I. Fletcher C, Peto R, Tinker CM. A comparison of the assessment of simple bronchitis (chronic mucus hypersecretion) by measurements of sputum volume and by standard questions on phlegm production, lnt J Epidemiol 1974; 3: 315-319.
Observations on sputum production
12. Brompton Hospital/Medical Research Council Collaborative trial: double blind trial comparing two dosage schedules of beclomethasone diproprionate aerosol in chronic bronchial asthma. Chest 1979; 73: 121. 13. Hendrick DJ, Davies RJ, D'Souza MF, Pepys J. An
31
analysis of skin prick tests in 656 asthmatic patients. Thorax 1975, 30: 2. 14. Peat JK, Woolcock AJ, Cullen K. Rate ofdecline of lung function in subjects with asthma. Eur J Respir Dis 1987, 70: 171-179.