Changes in Total Lung Capacity during Acute Spontaneous Asthma 1 - 3
S. P. BLACKIE,4 S. AL-MAJED, C. A. STAPLES, C. HILLlAM, and P. D. PARE
Introduction
Acute attacks of asthma result in dramatic changes in lung function. The consequences of airway obstruction include gas trapping (increase in residual volume) and hyperinflation (increasein functional residual capacity). It has been suggested that TLC increases acutely during severe attacks of asthma and then returns to normal, coincident with recovery from airway obstruction (1-5). Other investigators have failed to find any increase in TLC associated with acute airway narrowing (6-9). Since the earlier studies were published, it has been recognized that both helium dilution and plethysmography are subject to errors in the estimation of lung volumes in the presence of airway obstruction (7, 10), making interpretation of the findings of previous studies difficult. Chest roentgenograms have been used as an alternative method to measure TLC (11). Measurement ofTLC by radiographic techniques has the advantage of being independent of airway obstruction and has been used previously in studies of patients who have obstructive lung disease (12). We have investigated whether there is an increase in TLC during acute attacks of asthma by measuring radiographic TLC in patients admitted to hospital with acute spontaneous asthma. Methods Patient Population All asthmatic patients admitted to hospital by the medical service who were suffering from acute asthma were considered for entry into the study. Asthma was diagnosed on the basis of a history of episodic shortness of breath and wheezing without persistent symptoms between episodes. Treatment with nebulized ~2-agonists, intravenous aminophylline, or systemic corticosteroids was directed by the medical service and did not influence patient selection. Patients who were too ill to perform spirometry or in whom satisfactory posteroanterior (PA) and lateral chest radiographs could not be obtained were excluded. Twelve patients were entered into the study over a 6-month period.
SUMMARY An increased TLC has been reported during exacerbations of asthma, but the methods used (helium dilution, plethysmography) have been subsequently found unreliable In the assessment of lung volumes in patients with obstructive lung disease. To address this problem, we measured TLC (TLC-XR) from posteroanterior and lateral chest roentgenograms obtained during exacerbations (E) of asthma and after recovery (R) using planimetry in 12 asthmatic subjects. At recovery, TLC was also measured by plethysmography or by helium dilution for comparison with the radiographic measurement. The plethysmographic measurements were made with a panting frequency less than 1 Hz to allow for airway obstruction. A chest radiologist also used independent radiologic measurements of hyperinflation (lung height, diaphragmatic arc height, rib counts) to assess lung volumes. Mean FEV1 during E was 1.43 ± 0.38 L, and significant improvement occurred at R (FEV1 = 2.81 ± 0.58 L, P < 0.05). Of the independent radiologic variables measured, only an increase in lung height distinguished the two sets of radiographs. Mean TLC-XR(E) (6.01 ± 1.62 L) was significantly greater than mean TLC-XR (R) (5.44 ± 1.17 L, P < 0.05). TLC measured radiographically at recovery was strongly correlated (r = 0.94) with TLC measured by plethysmography or helium dilution. We conclude that acute reversible increases in TLC do occur during exacerbations of asthma and that these changes are only readily detected by formal planimetry. AM REV RESPIR DIS 1990; 142:79-83
Protocol Once a patient was identified as meeting the entry criteria, spirometry was performed using a linear dry wedge-type bellows (S-Model; Vitalograph, Buckingham, UK). Three reproducible efforts were obtained, and the highest FEV 1 value was recorded. Chest radiographs (PA and lateral) were obtained in the emergency radiology department. The patients performed repeat spirometry throughout their hospital course and, if possible, after discharge. Repeat chest radiographs corresponding as closely as possible to the time of their maximal FEV 1 were obtained. At the time of recovery, measurements of TLC (TLC at recovery = TLC-R) were also made using plethysmography or helium dilution. A pressure-compensated volume-displacement plethysmograph was used in 10patients. For each subject, we determined TLC-R by first measuring thoracic gas volume (Boyle's Law technique) followed by an inspiratory capacity maneuver. Alveolar pressure changes were estimated from mouth pressure changes measured using a differential transducer (MP 45-2, range ± 100 em H 20 at 100 kPa; Validyne Corp., Northridge, CAl. At least five measurements of thoracic gas volume and inspiratory pressure were obtained in each subject, and TLC-R was calculated by adding the mean thoracic gas volume to the mean inspiratory capacity. Volume was measured with a Krogh spirometer coupled to a linear displacement transducer (Type 300 HR; Shaevitz, Pennsauken, NJ). Patients were instructed to pant
slowly « 1 Hz) to minimize the overestimation of lung volume, which has been reported in the presence of airways obstruction (13). The remaining two patients had measurements of TLC made using a standard helium dilution technique (14) (Model C; P. K. Morgan, Chatham, Kent, UK). The helium tests were considered complete after 7 min or at that point when helium concentration did not change by more than 0.04070 in a 60-s interval. These tests were performed once for each subject.
Radiographic Determination of Lung Volume Roentgenographic TLC (TLC-XR) was measured from standard PA and lateral radiographs obtained during the spontaneous exacerbation and after recovery using the method described by Harris and colleagues (15).
(Received in original form August 10, 1989 and in revised form December lJ, 1989) I From the UBC Pulmonary Research Laboratory, St. Paul's Hospital, Vancouver, British Columbia, Canada. 2 Supported by the Lung Association of British Columbia. J Correspondence and requests for reprints should be addressed to Dr. S. P. Blackie, Pulmonary Research Laboratory, St. Paul's Hospital, 1081 Burrard St., Vancouver, BC, Canada V6Z lY6. 4 Recipient of a Fellowship from the Canadian Lung Association.
79
80
BLACKIE, AL-MAJED, STAPLES, HILLlAM, AND PARE
TABLE 1 PATIENT CHARACTERISTICS AND LUNG FUNCTION
Patient No.
1 2 3 4 5 6 7 8 9 10 11 12 Mean ± SO
Sex F M M F F F F M F F F F
FEV,
FVC
(L)
(L)
FEV,/FVC (%)
Age (yr)
Acute
Recovery
Acute
Recovery
Acute
Recovery
20 27 25 28 39 32 32 51 30 21 23 24 29 8.3
0.49 2.17 1.53 1.33 1.55 1.21 1.62 1.26 1.64 1.64 1.48 1.21 1.43 0.38
3.53 2.51 3.00 2.60 2.46 3.76 3.14 2.44 2.32 1.86 3.68 2.44 2.81 * 0.58
1.02 5.75 2.62 2.64 2.30 3.05 3.04 2.87 2.85 2.85 2.37 1.77 2.76 1.11
3.86 5.80 3.76 3.56 3.20 4.95 4.36 3.16 3.74 2.87 3.96 3.48 3.89* 0.82
48 38 58 50 67 40 53 44 58 58 62 68 54 10.1
91 43 80 73 77 76 72 77 62 65 93 70 73* 13.2
• Significantly different (p
< 0.05)
from corresponding acute value.
TABLE 2 TOTAL LUNG CAPACITY: ACUTE AND RECOVERY TlC-R (L)
TLC-XR Patient No.
1 2 3 4 5 6 7 8 9 10 11 12 Mean ± SO
(L)
Acute
Recovery
Change
Recovery
(% pred)
4.43 9.02 4.50 5.63 4.55 7.30 6.11 9.32 4.95 5.99 4.87 5.40 6.01 1.62
4.14 7.29 3.91 5.43 3.97 6.11 5.34 7.91 5.47 5.17 5.23 5.31 5.44* 1.17
-0.29 -1.73 -0.59 -0.20 -0.58 -1.19 -0.77 -1.41 0.52 -0.82 0.36 -0.09 -0.57 0.65
4.64 9.11 4.79 5.25 (He) 4.71 (He) 6.42 5.78 8.91 5.76 4.93 5.36 6.13 5.98t 1.52
85 109 77 99 94 106 105 132 122 89 93 104 101
Definition of abbreviations: TLC-XA = roentgenographic TLC; TLC-A mography or helium dilution) measured at recovery. • Significantly different (p < 0.05) from corresponding acute value. t Significantly different (p < 0.05) from TLC-XA at recovery.
Lung borders were traced using a digitizing board (Type 5 System; GTCO Corp, Rockville, MD) to determine lung area. Using the regression equation from the study of Harris and colleagues, the TLC-XR was calculated: TLC (ml) = 8.5 x roentgenographic lung area (cm-) - 1,200. Todetermine interobserver variability,measurements weremade bytwo observers who were blinded to patient identity and radiographic date. To determine intraobserver variability, a thoracic radiologist repeated the measurements on a separate occasion in a similarly blinded fashion using the same method to determine TLC. The measurements of interobserver and intraobserver variability weredone in only the first seven patients after it was noted that no variability was present. We also measured four radiographic indices (the height of the right lung in the posteroanterior projection, the height of the arc of the right diaphragm in the lateral projection, the anterior right rib count, and
= physiologic
TLC (plethys-
the posterior right rib count) to assess the degree of lung inflation (16).Using a combination of these four measurements, an observer blinded to the timing of each film tried to distinguish between the acute and recovery radiographs of each patient.
Data Analysis The values for FEV I ' TLC-XR, and the four independent radiographic variablesduring the acute attack were compared with those during recovery using paired t tests. Linear correlation coefficients for interobserver and intraobserver variability and for comparing the results of the different methods of measuring TLC (TLC-XR, TLC-R) were obtained using the least squares method.
Results Twelve patients were entered into the study. Patient characteristics and baseline
lung function are displayed in table 1.The average duration between measurements was 26.5 days. FEV 1 at the time of entry into the study was 1.43 ± 0.38 L (mean ± SD). A significant improvement in FEV 1 was seen by the time of recovery (2.81 ± 0.58 L, p < 0.05). FVC at entry was 2.76 ± 1.11 L. A significant improvement in FVC was seen at recovery (3.89 ± 0.82 L, p < 0.05). Mean TLC-XR at presentation was 6.01 ± 1.62 L (table 2). In 10 of 12 patients there was a decrease in TLC-XR at the time of recovery (figure 1). At the time of follow-up, mean TLC-XR had significantly decreased to 5.44 ± 1.17 L (p < 0.05). There was a strong correlation (r = 0.94) (figure 2) between TLC-XR at recovery and TLC-R (as measured by body plethysmograph or helium dilution), although there was a significant difference (p < 0.05) in the actual values obtained using the two methods (table 2). Interobserver variability (figure 3, upper panel) of the measurements of TLCXR made by CS and CH was minimal (r = 0.98). Intraobserver variability (figure 3, lower panel) between the two measurements of TLC-XR made at separate sittings by CS was negligible (r = 0.99). Of the independent radiologic variables measured, only lung height changed between the films taken during the acute phase and recovery. The mean decline in lung height was 0.43 ± 0.84 em (mean ± SD), which was statistically significant (p < 0.05). Rib counts and a formal measurement of diaphragmatic arc height were unable to distinguish between films in the individual patients. However, when all of the measurements were looked at together, we were able to correctly identify the acute and recovery films in 11 of 12 patients. No consistent relationship was found between the change in TLC-XR and the change in FEV I (figure 4). To further assess the factors affecting change in TLC, we divided the patients into three groups based on the interval between their two films. The three groups were (1) patients in whom follow-up films were obtained. less than 1 wk after the initial film (n = 4), (2) patients in whom follow-up was between 1 wk and 1 month (n = 3), and (3) patients in whom follow-up was greater than 1 month (n = 5). Average followup time in the three groups was 3 days, 25 days, and 111 days. The mean changes in FEV I did not appear to be related to time (1.27, 1.90, and 1.17 L, respectively), but the mean change in TLC increased as time between examinations in-
81
TOTAL WNG CAPACITY AND ACUTE ASTHMA
creased (+0.03, -0.76, and -0.92 L, respectively, 1 versus 3; p < 0.05).
10 9
.
Fig. 1. Change in TLC-XR from acute episode to recovery. Each line represents an individual patient. Means and their standard deviations are represented by the closed circles and the vertical bars. There is a significant difference between the mean TLC-XR of the acute episode and recovery.
8
.....
§
7
C)
6
,..l
~ I
~
•
j
fo-
I•
1
~
4
Acute
Recovery
10
a Fig. 2. Roentgenographic TLC (TLCXR) during recovery versus physiologic TLC (TLC-R) during recovery. The open circles represent TLC-R measured by plethysmography, and the crosses represent TLC-R measured by helium dilution. Solid line represents line of identity.
a
.... Ill:
a
>
S. P. BLACKIE,4 S. AL-MAJED, C. A. STAPLES, C. HILLlAM, and P. D. PARE
Introduction
Acute attacks of asthma result in dramatic changes in lung function. The consequences of airway obstruction include gas trapping (increase in residual volume) and hyperinflation (increasein functional residual capacity). It has been suggested that TLC increases acutely during severe attacks of asthma and then returns to normal, coincident with recovery from airway obstruction (1-5). Other investigators have failed to find any increase in TLC associated with acute airway narrowing (6-9). Since the earlier studies were published, it has been recognized that both helium dilution and plethysmography are subject to errors in the estimation of lung volumes in the presence of airway obstruction (7, 10), making interpretation of the findings of previous studies difficult. Chest roentgenograms have been used as an alternative method to measure TLC (11). Measurement ofTLC by radiographic techniques has the advantage of being independent of airway obstruction and has been used previously in studies of patients who have obstructive lung disease (12). We have investigated whether there is an increase in TLC during acute attacks of asthma by measuring radiographic TLC in patients admitted to hospital with acute spontaneous asthma. Methods Patient Population All asthmatic patients admitted to hospital by the medical service who were suffering from acute asthma were considered for entry into the study. Asthma was diagnosed on the basis of a history of episodic shortness of breath and wheezing without persistent symptoms between episodes. Treatment with nebulized ~2-agonists, intravenous aminophylline, or systemic corticosteroids was directed by the medical service and did not influence patient selection. Patients who were too ill to perform spirometry or in whom satisfactory posteroanterior (PA) and lateral chest radiographs could not be obtained were excluded. Twelve patients were entered into the study over a 6-month period.
SUMMARY An increased TLC has been reported during exacerbations of asthma, but the methods used (helium dilution, plethysmography) have been subsequently found unreliable In the assessment of lung volumes in patients with obstructive lung disease. To address this problem, we measured TLC (TLC-XR) from posteroanterior and lateral chest roentgenograms obtained during exacerbations (E) of asthma and after recovery (R) using planimetry in 12 asthmatic subjects. At recovery, TLC was also measured by plethysmography or by helium dilution for comparison with the radiographic measurement. The plethysmographic measurements were made with a panting frequency less than 1 Hz to allow for airway obstruction. A chest radiologist also used independent radiologic measurements of hyperinflation (lung height, diaphragmatic arc height, rib counts) to assess lung volumes. Mean FEV1 during E was 1.43 ± 0.38 L, and significant improvement occurred at R (FEV1 = 2.81 ± 0.58 L, P < 0.05). Of the independent radiologic variables measured, only an increase in lung height distinguished the two sets of radiographs. Mean TLC-XR(E) (6.01 ± 1.62 L) was significantly greater than mean TLC-XR (R) (5.44 ± 1.17 L, P < 0.05). TLC measured radiographically at recovery was strongly correlated (r = 0.94) with TLC measured by plethysmography or helium dilution. We conclude that acute reversible increases in TLC do occur during exacerbations of asthma and that these changes are only readily detected by formal planimetry. AM REV RESPIR DIS 1990; 142:79-83
Protocol Once a patient was identified as meeting the entry criteria, spirometry was performed using a linear dry wedge-type bellows (S-Model; Vitalograph, Buckingham, UK). Three reproducible efforts were obtained, and the highest FEV 1 value was recorded. Chest radiographs (PA and lateral) were obtained in the emergency radiology department. The patients performed repeat spirometry throughout their hospital course and, if possible, after discharge. Repeat chest radiographs corresponding as closely as possible to the time of their maximal FEV 1 were obtained. At the time of recovery, measurements of TLC (TLC at recovery = TLC-R) were also made using plethysmography or helium dilution. A pressure-compensated volume-displacement plethysmograph was used in 10patients. For each subject, we determined TLC-R by first measuring thoracic gas volume (Boyle's Law technique) followed by an inspiratory capacity maneuver. Alveolar pressure changes were estimated from mouth pressure changes measured using a differential transducer (MP 45-2, range ± 100 em H 20 at 100 kPa; Validyne Corp., Northridge, CAl. At least five measurements of thoracic gas volume and inspiratory pressure were obtained in each subject, and TLC-R was calculated by adding the mean thoracic gas volume to the mean inspiratory capacity. Volume was measured with a Krogh spirometer coupled to a linear displacement transducer (Type 300 HR; Shaevitz, Pennsauken, NJ). Patients were instructed to pant
slowly « 1 Hz) to minimize the overestimation of lung volume, which has been reported in the presence of airways obstruction (13). The remaining two patients had measurements of TLC made using a standard helium dilution technique (14) (Model C; P. K. Morgan, Chatham, Kent, UK). The helium tests were considered complete after 7 min or at that point when helium concentration did not change by more than 0.04070 in a 60-s interval. These tests were performed once for each subject.
Radiographic Determination of Lung Volume Roentgenographic TLC (TLC-XR) was measured from standard PA and lateral radiographs obtained during the spontaneous exacerbation and after recovery using the method described by Harris and colleagues (15).
(Received in original form August 10, 1989 and in revised form December lJ, 1989) I From the UBC Pulmonary Research Laboratory, St. Paul's Hospital, Vancouver, British Columbia, Canada. 2 Supported by the Lung Association of British Columbia. J Correspondence and requests for reprints should be addressed to Dr. S. P. Blackie, Pulmonary Research Laboratory, St. Paul's Hospital, 1081 Burrard St., Vancouver, BC, Canada V6Z lY6. 4 Recipient of a Fellowship from the Canadian Lung Association.
79
80
BLACKIE, AL-MAJED, STAPLES, HILLlAM, AND PARE
TABLE 1 PATIENT CHARACTERISTICS AND LUNG FUNCTION
Patient No.
1 2 3 4 5 6 7 8 9 10 11 12 Mean ± SO
Sex F M M F F F F M F F F F
FEV,
FVC
(L)
(L)
FEV,/FVC (%)
Age (yr)
Acute
Recovery
Acute
Recovery
Acute
Recovery
20 27 25 28 39 32 32 51 30 21 23 24 29 8.3
0.49 2.17 1.53 1.33 1.55 1.21 1.62 1.26 1.64 1.64 1.48 1.21 1.43 0.38
3.53 2.51 3.00 2.60 2.46 3.76 3.14 2.44 2.32 1.86 3.68 2.44 2.81 * 0.58
1.02 5.75 2.62 2.64 2.30 3.05 3.04 2.87 2.85 2.85 2.37 1.77 2.76 1.11
3.86 5.80 3.76 3.56 3.20 4.95 4.36 3.16 3.74 2.87 3.96 3.48 3.89* 0.82
48 38 58 50 67 40 53 44 58 58 62 68 54 10.1
91 43 80 73 77 76 72 77 62 65 93 70 73* 13.2
• Significantly different (p
< 0.05)
from corresponding acute value.
TABLE 2 TOTAL LUNG CAPACITY: ACUTE AND RECOVERY TlC-R (L)
TLC-XR Patient No.
1 2 3 4 5 6 7 8 9 10 11 12 Mean ± SO
(L)
Acute
Recovery
Change
Recovery
(% pred)
4.43 9.02 4.50 5.63 4.55 7.30 6.11 9.32 4.95 5.99 4.87 5.40 6.01 1.62
4.14 7.29 3.91 5.43 3.97 6.11 5.34 7.91 5.47 5.17 5.23 5.31 5.44* 1.17
-0.29 -1.73 -0.59 -0.20 -0.58 -1.19 -0.77 -1.41 0.52 -0.82 0.36 -0.09 -0.57 0.65
4.64 9.11 4.79 5.25 (He) 4.71 (He) 6.42 5.78 8.91 5.76 4.93 5.36 6.13 5.98t 1.52
85 109 77 99 94 106 105 132 122 89 93 104 101
Definition of abbreviations: TLC-XA = roentgenographic TLC; TLC-A mography or helium dilution) measured at recovery. • Significantly different (p < 0.05) from corresponding acute value. t Significantly different (p < 0.05) from TLC-XA at recovery.
Lung borders were traced using a digitizing board (Type 5 System; GTCO Corp, Rockville, MD) to determine lung area. Using the regression equation from the study of Harris and colleagues, the TLC-XR was calculated: TLC (ml) = 8.5 x roentgenographic lung area (cm-) - 1,200. Todetermine interobserver variability,measurements weremade bytwo observers who were blinded to patient identity and radiographic date. To determine intraobserver variability, a thoracic radiologist repeated the measurements on a separate occasion in a similarly blinded fashion using the same method to determine TLC. The measurements of interobserver and intraobserver variability weredone in only the first seven patients after it was noted that no variability was present. We also measured four radiographic indices (the height of the right lung in the posteroanterior projection, the height of the arc of the right diaphragm in the lateral projection, the anterior right rib count, and
= physiologic
TLC (plethys-
the posterior right rib count) to assess the degree of lung inflation (16).Using a combination of these four measurements, an observer blinded to the timing of each film tried to distinguish between the acute and recovery radiographs of each patient.
Data Analysis The values for FEV I ' TLC-XR, and the four independent radiographic variablesduring the acute attack were compared with those during recovery using paired t tests. Linear correlation coefficients for interobserver and intraobserver variability and for comparing the results of the different methods of measuring TLC (TLC-XR, TLC-R) were obtained using the least squares method.
Results Twelve patients were entered into the study. Patient characteristics and baseline
lung function are displayed in table 1.The average duration between measurements was 26.5 days. FEV 1 at the time of entry into the study was 1.43 ± 0.38 L (mean ± SD). A significant improvement in FEV 1 was seen by the time of recovery (2.81 ± 0.58 L, p < 0.05). FVC at entry was 2.76 ± 1.11 L. A significant improvement in FVC was seen at recovery (3.89 ± 0.82 L, p < 0.05). Mean TLC-XR at presentation was 6.01 ± 1.62 L (table 2). In 10 of 12 patients there was a decrease in TLC-XR at the time of recovery (figure 1). At the time of follow-up, mean TLC-XR had significantly decreased to 5.44 ± 1.17 L (p < 0.05). There was a strong correlation (r = 0.94) (figure 2) between TLC-XR at recovery and TLC-R (as measured by body plethysmograph or helium dilution), although there was a significant difference (p < 0.05) in the actual values obtained using the two methods (table 2). Interobserver variability (figure 3, upper panel) of the measurements of TLCXR made by CS and CH was minimal (r = 0.98). Intraobserver variability (figure 3, lower panel) between the two measurements of TLC-XR made at separate sittings by CS was negligible (r = 0.99). Of the independent radiologic variables measured, only lung height changed between the films taken during the acute phase and recovery. The mean decline in lung height was 0.43 ± 0.84 em (mean ± SD), which was statistically significant (p < 0.05). Rib counts and a formal measurement of diaphragmatic arc height were unable to distinguish between films in the individual patients. However, when all of the measurements were looked at together, we were able to correctly identify the acute and recovery films in 11 of 12 patients. No consistent relationship was found between the change in TLC-XR and the change in FEV I (figure 4). To further assess the factors affecting change in TLC, we divided the patients into three groups based on the interval between their two films. The three groups were (1) patients in whom follow-up films were obtained. less than 1 wk after the initial film (n = 4), (2) patients in whom follow-up was between 1 wk and 1 month (n = 3), and (3) patients in whom follow-up was greater than 1 month (n = 5). Average followup time in the three groups was 3 days, 25 days, and 111 days. The mean changes in FEV I did not appear to be related to time (1.27, 1.90, and 1.17 L, respectively), but the mean change in TLC increased as time between examinations in-
81
TOTAL WNG CAPACITY AND ACUTE ASTHMA
creased (+0.03, -0.76, and -0.92 L, respectively, 1 versus 3; p < 0.05).
10 9
.
Fig. 1. Change in TLC-XR from acute episode to recovery. Each line represents an individual patient. Means and their standard deviations are represented by the closed circles and the vertical bars. There is a significant difference between the mean TLC-XR of the acute episode and recovery.
8
.....
§
7
C)
6
,..l
~ I
~
•
j
fo-
I•
1
~
4
Acute
Recovery
10
a Fig. 2. Roentgenographic TLC (TLCXR) during recovery versus physiologic TLC (TLC-R) during recovery. The open circles represent TLC-R measured by plethysmography, and the crosses represent TLC-R measured by helium dilution. Solid line represents line of identity.
a
.... Ill:
a
>
