Pulmonary Function in Human Immunodeficiency Virus Infection A Prospective 18-Month Study of Serial Lung Function in 474 Patients 1 - 4
D. M. MITCHELL, J. FLEMING, A. J. PINCHING, J. R. W. HARRIS, F. M. MOSS, D. VEALE, and R. J. SHAW
The
Introduction
reduction in carbon monoxide diffusing capacity (Oleo) in patients with Pneumocystis carinii pneumonia (PCP) is well recognized (1, 2).-However, the clinical usefulness of identifying a low OLeo is limited: in a previous study (3) wefound evidenceof a reduction in OleO in human immunodeficiency virus (HIV)-infected patients who did not have PCP. The reason for this reduction in OLeo in HIV-related disease is unknown. To explain the phenomenon we generated two hypotheses. First, since HIV can be identified in the lungs of patients with acquired immunodeficiency syndrome (AIOS) (4,5), wehypothesized that HIV infection per se caused a subtle lung injury that manifested itself by a gradual decline in OLeo. In the second alternative hypothesis, we considered that HIV had no direct action but that HIVinduced immunosuppression impaired the ability of the lung to recover from inflammatory events. To test these hypotheses we studied serial lung function in 474 HIV-seropositive patients. We found no evidence of a gradual decline in OLeo, nor was administration of the antiviral agent zidovudine (AZT) associated with an improved OLeo. Following the single inflammatory stimulus of PCP, there was a slow improvement in OLeo over 3 months that did not achieve predicted values. When the second inflammatory stimulus of smoking was also present, the extent of the recovery in OLeo after PCP was further limited. Similarly, the inflammatory stimulus of smoking itself was associated with a reduction in Dtoo irrespective of disease category. Methods A total of 474 male homosexual or bisexual HIV-seropositive patients who presented to the sexually transmitted disease (STD) clinic
SUMMARY To Investigate the development of a reduced Oleo In patients with HIV-related disease, we studied 474 HIV-seroposltlve patients and performed serial lung function measurements over 18 months. The mean values of OLeo at presentation were lower In patients whh more advanced HIV disease compared whh asymptomatic HIV-seroposltlve patients (OLeo 88% of predicted). When compared with the Oleo In asymptomatic HIV-seroposltlve patients, the OLeo had reduced values In patients with persistent generalized lymphadenopathy (PGL) (82% of predicted, P < 0.05), acquired deficiency syndrome-related complex (ARC) (73% predicted, p< 0.001),non pUlmonary Kaposi's sarcoma (KS) (72% of predicted, p < 0.001), nonpulmonary complications of AIDS excluding KS (73% of predicted, p < 0.001), pulmonary KS (63% of predicted, P < 0.001),pulmonary mycobacterlallntectlon (68% of predicted, p < 0.05), pyogenic Infection (70%, p < 0.05), acute Pneumocystls ca,'nllpneumonia (PCPj 49%, P < 0.001), and following recovery from PCP (71%, P < 0.001).Serial lung function measurements over 18 months revealed no change In Oleo within any patient group, and In particular there was no tendency for a gradual decline. Clinical deterioration due to the development of PCP was associated with a reduction In Oleo. Conversely, In patients recovering from PCp, there was a partial Improvement In OLeo over 3 months. Zidovudine (AZT) use did not affect OLeo within any diagnostic group or the recovery In Oleo following PCP. However, cigarette smokIng was associated with further reductions In OLeo In all patient groups and with an Impaired recovery of OLeo following acute PCP.This stUdy confirmed that OleO was reduced In patients with pulmonary and non pulmonary HIV-related complications. There was no evidence of an Insidious deAM REV RESPIR DIS 1992; 146:745-751 cline In OLeo or of a beneficial effect of zldovudlne.
were studied. A drug history was taken from all subjects, and all denied intravenous drug abuse. Consecutive patients who attended the sexually transmitted disease clinic from 1985 wereenrolled. The criteria for inclusion in the study were (1) positive HIV serology, (2) no history of intravenous drug abuse, and (3) informed consent. Exclusion criteria were (1) respiratory failure and (2) refusal of the patient to participate. The study group included a small proportion of individuals previously known to the clinic, as well as a larger group of new patients referred on the basis of symptoms or for HIV testing. The patients were classified according to clinical status in line with Centers for Disease Control (CDC) criteria at the time of the first lung function measurements (table 1) (6): asymptomatic HIV-seropositive (HIV+), persistent generalized lymphadenopathy (POL), AIDS-related complex (ARC), nonpulmonary Kaposi's sarcoma (KS; mainly skin Kaposi's sarcoma), nonpulmonary, non-KS AIDS (opportunistic ;.nfection involving other organs), AIDS patients with pulmonary complications (pulmonary AIDS), acute PCP, patients 3months following recovery from PCP (post PCP),
lung mycobacterial infection (lung mycobacteria), lung Kaposi's sarcoma (lung KS), and AIDS patients with pulmonary pyogenicbacterial infection. To identify concurrent inflammatory stimuli to the lung, patients wereclassified as smokers (current smokers or patients who had smoked within the previous 6 months) or nonsmokers (patients who had never smoked or abstained for at least 6 months). In addition, patients were classified according to whether they werereceivingregular prophylaxis with AZT (200to 500mg/day) (Wellcome, Crewe, UK). (Received in original form April 25, 1991 and in revised form March 16, 1992) 1 From the Departments of Respiratory and Genitourinary Medicine, St. Mary's Hospital, london, United Kingdom. 2 Supported by the British Lung Foundation. 3 Presented in part at the Annual Meeting of the American Thoracic Society, Las Vegas,May 1988. 4 Correspondence and requests for reprints should be addressed to Dr. D. M. Mitchell, Chest and Allergy Clinic, St. Mary's Hospital, Praed Street, London, W2 INY, UK.
745
746
MITCHELL, FLEMING, PINCHING, HARRIS, MOSS, VEALE, AND SHAW
TABLE 1 CLINICAL CHARACTERISTICS AND CDC CRITERIA FOR EACH GROUp·
CDC Group
Group
Clinical Classification
HIV+ PGL ARC
Asymptomatic HIV seropositive Persistent generalized lymphadenopathy AIDS-related complex
KS
Nonpulmonary Kaposi's sarcoma (KS), mainly skin KS Nonpulmonary AIDS without KS: opportunistic infection of other organs Pulmonary complications of AIDS
Non-KS AIDS
Pulmonary AIDS, including Acute PCP Acute P. carinii pneumonia (PCP) Post PCP Lung KS Lung myco Lung pyog
II 11\ IV subgroup A, B, C z, E IV subgroup D IV subgroup C1
IV subgroup C,
3 Months following recovery from PCP Lung Kaposi's sarcoma Lung mycobacterial infection AIDS patients with pyogenic bacterial lung infections
(yr)
Smoker (% of patients)
Mild Dyspnea (% of patients)
126 95
20-58 23-54
56 59
24 31
8 6
106 102
96 90
92
19-59
57
39
6
96
87
35
23-57
34
6
11
95
86
19 107
22-50
40
30
5
98
86
78 42 11 11 7
24-60 24-60 26-53 25-60 29-44
34
53 36 27 47 66
27 9 9 20 0
78 95 87 91 101
74 84 88 79 84
Number of Patients
Age Range
31
45 63 29
Severe SOB FEV, (% of (% of patients) predicted)
PEFR (% of predicted)
• Number of patients, age, smoking habit, prevalence of dyspnea, FEVlo and peak flow in each clinical category. Clinical categories identified by group name and CDC grouping.
Source: From reference 9.
Patients seropositive for HIV with an abnormal chest radiograph or DLco of less than 70070 predicted with .espiratory symptoms (persistent cough or moderate to severe exertional dyspnea) underwent diagnostic fiberoptic bronchoscopy with bronchoalveolar lavage with or without transbronchial biopsy (7). Cellular material in the lavage was collected by centrifugation and stained using Papanicolaou, Grocott, and May-Griinwald-Giemsa stains before microscopic examination (8). Biopsies were fixed and stained with hematoxylin and eosin, Grocott, periodic acidSchiff, and Ziel-Nielson stains. All patients had the following lung function tests: forced expiratory volume in one second (FEV 1), peak expiratory flow (PEF), forced vital capacity (FVC), alveolar volume (VA), and DLeo. The diffusion coefficient (Keo) was derived from the DLeo divided by the VA. Measurements weremade using a dry bellows spirometer and the single-breath helium dilution method (pK Morgan transfer test model C machine; P. K. Morgan, Gillingham, Kent). Corrections for temperature and hemoglobin (9) wereincluded when calculating values for OLeo. To ensure accuracy overtime, the same machine was used for all tests and the machine was calibrated daily using standard gases. All measurements were made after a period of rest. The lung function tests were repeated at 3-month intervals in all patients for the duration of the study. Additional measurements were performed in patients who developed respiratory symptoms or in those with clinicaldeterioration. Each test was performed three times, and the best value was recorded. To reduce the risk of cross-infection between patients, modifications were made to the equipment. Lung function testing apparatus designated for HIV-seropositive patients
was used throughout. One-way valve safety mouthpieces (Vitalograph Ltd., Buckingham) were used for spirometry to avoid the risk of inhaling pathogens from the spirometer. A Pall Ultipors breathing system filter (pall Biomedical, Havant, Portsmouth) was placed distal to the mouthpiece during the CO transfer test. The increased tidal volume of 150ml due to the presence of the filter was incorporated in the calculation. Plastic mouthpieces and nose clips were sterilized in 2% glutaraldehyde for 1 h between patients and the remainder of the apparatus dismantled and sterilized weekly for >3 h in 2% glutaraldehyde. For each patient, values of each lung function measurement were compared with those predicted for age, sex, and height (9). The results, expressed as a percentage of predicted for each patient, were grouped according to the clinical classification of patients, and statistical comparisons were made using the Mann-Whitney U test, the Wilcoxon matched-pairs signed-rank test, or analysis of variance (ANOVA) as indicated. Data are presented as mean ± standard error of the mean (SEM). The study was approved by he hospital ethics committee.
Results
Pulmonary hmction on Presentation A group of 4i4 patients werestudied. The number with each clinical diagnosis is shown in table 1. The age, percentage who were still smoking at the time of the study, and percentage who had mild or severe dyspnea in each patient group are listed in table 1. Nonsmokers were defined as those who had never smoked or those who had abstained from smoking for> 6 months. Mild dyspnea was record-
ed if the patient complained of breathlessness on stairs or hills, severe dyspnea if breathless on minimal exertion or at rest. Patients with acute PCP and lung mycobacterial disease had a high incidence of dyspnea. In the remaining groups up to 300,10 had mild dyspnea and approximately 10070 of all groups complained of severe dyspnea. Airway function (FEV 1 and PEF) was normal in most groups of HIV-infected patients. Patients with acute PCP had reduced values of FEV. and PEFR, and there were small reductions in patients with pulmonary KS (table 1). The mean result of pulmonary function tests on asymptomatic HIV-positive patients at the first visit revealed slight reductions in DLeo (880,10 of predicted) and Keo (92070 of predicted), but not FVC compared with predicted normal for the British population according to age, sex, height, and ethnic group (figure 1) (9). All other groups of patients were classified clinically at the time of the first lung function tests, and the results of the lung function measurements were compared with those from HIV-seropositiveasymptomatic patients. Patients with POL had a significant reduction in Keo (p < 0.01) and DLeo (820,10 of predicted: p < 0.05), and patients with ARC had a significant reduction in DLeo (730,10 of predicted), Keo, and FVC (p < 0.001, p < 0.001, and p < 0.001). Similarly, there were reductions in DLeo, Keo, and FVC in patients with nonpulmonary complications of AIDS. The DLeo (720,10 of predicted),
747
SERIAL PULMONARY FUNCTION IN HIV
100
T CO L
reductions in DLeo were also seen in the other categories of pulmonary disease complicating AIDS (pulmonary KS, 63070 of predicted, p< 0.001;mycobacterial infection, 68070 of predicted, p < 0.05; and pyogenic bacterial infection, 70070 of predicted, p < 0.05). The results of lung function tests on the presentation of these 474 patients include results on 169 patients that have been previously presented (3).
90 80 70 60 50 40
Keo 100 90 Q
W ~
o Ci W a: a.. 0~
80 70 60 50 40
120 110 100 90 80 70 60 50 40
HIV
PGL
ARC
+ve
n=
126
95
92
NON NON ACUTE POST LUNG LUNG LUNG PCP PCP KS KS AIDS 35
19
78
42
11
LUNG LUNG MYCO PYOG 11
7
Fig. 1. Lung function on presentation. OLeo. Keo, and FVC expressed as percentage of predicted in patients on entry into the study and classified according to clinical disease category (mean ± SEM). (***p < 0.001; **p < 0.Q1; *p < 0.05, Mann-Whitney U test.)
Kco, and FVC were reduced in AIDS patients with Kaposi's sarcoma but without pulmonary involvement (mainly cutaneous KS) (p < 0.001, p < 0.001, and p < 0.001)and in AIDS patients without KS and without pulmonary disease (opportunistic infections involving other organs), the DLeo (73070 of predicted), Keo,
and FVC were also reduced (p < 0.001, p < 0.001, and p < 0.05). In patients with AIDS and pulmonary complications, further reductions in lung function were 0 bserved. In patients with acute PCP marked reductions in mean Dteo, Keo, and FVC were found (p < 0.001, p < 0.001, and p < 0.001).Marked
Serial Lung Function over 18 Months in Clinically Stable Patients To dissociate subtle changes in clinically stable patients from changes associated with acute clinical deterioration, the results from serial lung function tests performed on the patients who did not change diagnostic group were analyzed separately. Pulmonary function tests were repeated at intervals of 3 months in patients from each category of HIV disease. Data were included from an 18month follow-up period or until the patient changed clinical category. In patients who did not change clinical diagnostic group, no significant reductions in DLeo, Keo, or FVC occurred in any patient group with the exception of pulmonary KS and post PCP (figure 2A). In patients with pulmonary KS there was some evidence of a gradual decline in FVC and DLeo. Following PCP there was a gradual improvement in DLeo (see later). Toensure that changes in individual patients were not obscured by grouping the data, the initial and 18-month values for individual patients who were followed throughout the study period and who did not change HIV category are presented separately (figure 2B). When compared by analysis of variance these individuals showed no tendency for a decline in the DLeo, with the exception of those individuals with lung KS. Lung Function in Patients with Clinical Deterioration Using the measurements from the 3month serial lung function tests, the results before and immediately following the deterioration were compared. Of the patients who deteriorated 41deteriorated such that they changed clinical diagnostic category during the period of the study. Of these patients, 8 who were initially asymptomatic HIV-seropositive patients or who had POL developed ARC. During this transition, there was no significant deterioration in DLeo or Keo; however, FVC results were more variable, with an increase in the four patients with POL who developed ARC
20
40
60
80
100
120
140
+~e I
6
9
12
15
18
23
16
14
91
I
8
15
7
15
18
101
18
17
12
I
18
120
20
40
o
3
6
MONTHS
20 I
40
60
15
4
60
12
5
80
~I
8
80
NON LUNG KS
In=35
100
~
140
,
3
o
43
i
3
o
25
6
ii,
39
6
6
6
,
o 3
6
i i i
7
3
i i
1n=11
20I
40
100
120
140
:11~
'
ln= 78
0
1N~N LUN~AI~I
3
140.-----In= 11 3 3 120
60
12
8
12
60
9
16
9
80
6
32
6
80
3
32
3
l
20 1
40
60
80
100
120
140
100
0
n= 87
0
IPGL
I
28
~
33
3
100
120
0
:1
60
9
6
4
15
i
12
5
12
,
3
15
3
15
,
3
18
12
I
1
15
i i i
7
,
3
I
12
2
i i '
14
9
i
3
I
120
140
20
40
60
80
100
120
140
80
~
20
40
~ 60
80
100
120
J
1
140
40
'#. 60
a..
~
~ 100
t> o
B
0
0
0
0
ARC
0
-
PGL
lr--
0
~
HIV+ve
-
18
18
I
--..,)01
....
18
Q
0
o
og....
LUNG KS
o
0-
~o
18
i
-0
0 . . . . - - - - - - -o
NON LUNG KS
40· 20
0
0
0
MONTHS
18
§=;:
D. M. MITCHELL, J. FLEMING, A. J. PINCHING, J. R. W. HARRIS, F. M. MOSS, D. VEALE, and R. J. SHAW
The
Introduction
reduction in carbon monoxide diffusing capacity (Oleo) in patients with Pneumocystis carinii pneumonia (PCP) is well recognized (1, 2).-However, the clinical usefulness of identifying a low OLeo is limited: in a previous study (3) wefound evidenceof a reduction in OleO in human immunodeficiency virus (HIV)-infected patients who did not have PCP. The reason for this reduction in OLeo in HIV-related disease is unknown. To explain the phenomenon we generated two hypotheses. First, since HIV can be identified in the lungs of patients with acquired immunodeficiency syndrome (AIOS) (4,5), wehypothesized that HIV infection per se caused a subtle lung injury that manifested itself by a gradual decline in OLeo. In the second alternative hypothesis, we considered that HIV had no direct action but that HIVinduced immunosuppression impaired the ability of the lung to recover from inflammatory events. To test these hypotheses we studied serial lung function in 474 HIV-seropositive patients. We found no evidence of a gradual decline in OLeo, nor was administration of the antiviral agent zidovudine (AZT) associated with an improved OLeo. Following the single inflammatory stimulus of PCP, there was a slow improvement in OLeo over 3 months that did not achieve predicted values. When the second inflammatory stimulus of smoking was also present, the extent of the recovery in OLeo after PCP was further limited. Similarly, the inflammatory stimulus of smoking itself was associated with a reduction in Dtoo irrespective of disease category. Methods A total of 474 male homosexual or bisexual HIV-seropositive patients who presented to the sexually transmitted disease (STD) clinic
SUMMARY To Investigate the development of a reduced Oleo In patients with HIV-related disease, we studied 474 HIV-seroposltlve patients and performed serial lung function measurements over 18 months. The mean values of OLeo at presentation were lower In patients whh more advanced HIV disease compared whh asymptomatic HIV-seroposltlve patients (OLeo 88% of predicted). When compared with the Oleo In asymptomatic HIV-seroposltlve patients, the OLeo had reduced values In patients with persistent generalized lymphadenopathy (PGL) (82% of predicted, P < 0.05), acquired deficiency syndrome-related complex (ARC) (73% predicted, p< 0.001),non pUlmonary Kaposi's sarcoma (KS) (72% of predicted, p < 0.001), nonpulmonary complications of AIDS excluding KS (73% of predicted, p < 0.001), pulmonary KS (63% of predicted, P < 0.001),pulmonary mycobacterlallntectlon (68% of predicted, p < 0.05), pyogenic Infection (70%, p < 0.05), acute Pneumocystls ca,'nllpneumonia (PCPj 49%, P < 0.001), and following recovery from PCP (71%, P < 0.001).Serial lung function measurements over 18 months revealed no change In Oleo within any patient group, and In particular there was no tendency for a gradual decline. Clinical deterioration due to the development of PCP was associated with a reduction In Oleo. Conversely, In patients recovering from PCp, there was a partial Improvement In OLeo over 3 months. Zidovudine (AZT) use did not affect OLeo within any diagnostic group or the recovery In Oleo following PCP. However, cigarette smokIng was associated with further reductions In OLeo In all patient groups and with an Impaired recovery of OLeo following acute PCP.This stUdy confirmed that OleO was reduced In patients with pulmonary and non pulmonary HIV-related complications. There was no evidence of an Insidious deAM REV RESPIR DIS 1992; 146:745-751 cline In OLeo or of a beneficial effect of zldovudlne.
were studied. A drug history was taken from all subjects, and all denied intravenous drug abuse. Consecutive patients who attended the sexually transmitted disease clinic from 1985 wereenrolled. The criteria for inclusion in the study were (1) positive HIV serology, (2) no history of intravenous drug abuse, and (3) informed consent. Exclusion criteria were (1) respiratory failure and (2) refusal of the patient to participate. The study group included a small proportion of individuals previously known to the clinic, as well as a larger group of new patients referred on the basis of symptoms or for HIV testing. The patients were classified according to clinical status in line with Centers for Disease Control (CDC) criteria at the time of the first lung function measurements (table 1) (6): asymptomatic HIV-seropositive (HIV+), persistent generalized lymphadenopathy (POL), AIDS-related complex (ARC), nonpulmonary Kaposi's sarcoma (KS; mainly skin Kaposi's sarcoma), nonpulmonary, non-KS AIDS (opportunistic ;.nfection involving other organs), AIDS patients with pulmonary complications (pulmonary AIDS), acute PCP, patients 3months following recovery from PCP (post PCP),
lung mycobacterial infection (lung mycobacteria), lung Kaposi's sarcoma (lung KS), and AIDS patients with pulmonary pyogenicbacterial infection. To identify concurrent inflammatory stimuli to the lung, patients wereclassified as smokers (current smokers or patients who had smoked within the previous 6 months) or nonsmokers (patients who had never smoked or abstained for at least 6 months). In addition, patients were classified according to whether they werereceivingregular prophylaxis with AZT (200to 500mg/day) (Wellcome, Crewe, UK). (Received in original form April 25, 1991 and in revised form March 16, 1992) 1 From the Departments of Respiratory and Genitourinary Medicine, St. Mary's Hospital, london, United Kingdom. 2 Supported by the British Lung Foundation. 3 Presented in part at the Annual Meeting of the American Thoracic Society, Las Vegas,May 1988. 4 Correspondence and requests for reprints should be addressed to Dr. D. M. Mitchell, Chest and Allergy Clinic, St. Mary's Hospital, Praed Street, London, W2 INY, UK.
745
746
MITCHELL, FLEMING, PINCHING, HARRIS, MOSS, VEALE, AND SHAW
TABLE 1 CLINICAL CHARACTERISTICS AND CDC CRITERIA FOR EACH GROUp·
CDC Group
Group
Clinical Classification
HIV+ PGL ARC
Asymptomatic HIV seropositive Persistent generalized lymphadenopathy AIDS-related complex
KS
Nonpulmonary Kaposi's sarcoma (KS), mainly skin KS Nonpulmonary AIDS without KS: opportunistic infection of other organs Pulmonary complications of AIDS
Non-KS AIDS
Pulmonary AIDS, including Acute PCP Acute P. carinii pneumonia (PCP) Post PCP Lung KS Lung myco Lung pyog
II 11\ IV subgroup A, B, C z, E IV subgroup D IV subgroup C1
IV subgroup C,
3 Months following recovery from PCP Lung Kaposi's sarcoma Lung mycobacterial infection AIDS patients with pyogenic bacterial lung infections
(yr)
Smoker (% of patients)
Mild Dyspnea (% of patients)
126 95
20-58 23-54
56 59
24 31
8 6
106 102
96 90
92
19-59
57
39
6
96
87
35
23-57
34
6
11
95
86
19 107
22-50
40
30
5
98
86
78 42 11 11 7
24-60 24-60 26-53 25-60 29-44
34
53 36 27 47 66
27 9 9 20 0
78 95 87 91 101
74 84 88 79 84
Number of Patients
Age Range
31
45 63 29
Severe SOB FEV, (% of (% of patients) predicted)
PEFR (% of predicted)
• Number of patients, age, smoking habit, prevalence of dyspnea, FEVlo and peak flow in each clinical category. Clinical categories identified by group name and CDC grouping.
Source: From reference 9.
Patients seropositive for HIV with an abnormal chest radiograph or DLco of less than 70070 predicted with .espiratory symptoms (persistent cough or moderate to severe exertional dyspnea) underwent diagnostic fiberoptic bronchoscopy with bronchoalveolar lavage with or without transbronchial biopsy (7). Cellular material in the lavage was collected by centrifugation and stained using Papanicolaou, Grocott, and May-Griinwald-Giemsa stains before microscopic examination (8). Biopsies were fixed and stained with hematoxylin and eosin, Grocott, periodic acidSchiff, and Ziel-Nielson stains. All patients had the following lung function tests: forced expiratory volume in one second (FEV 1), peak expiratory flow (PEF), forced vital capacity (FVC), alveolar volume (VA), and DLeo. The diffusion coefficient (Keo) was derived from the DLeo divided by the VA. Measurements weremade using a dry bellows spirometer and the single-breath helium dilution method (pK Morgan transfer test model C machine; P. K. Morgan, Gillingham, Kent). Corrections for temperature and hemoglobin (9) wereincluded when calculating values for OLeo. To ensure accuracy overtime, the same machine was used for all tests and the machine was calibrated daily using standard gases. All measurements were made after a period of rest. The lung function tests were repeated at 3-month intervals in all patients for the duration of the study. Additional measurements were performed in patients who developed respiratory symptoms or in those with clinicaldeterioration. Each test was performed three times, and the best value was recorded. To reduce the risk of cross-infection between patients, modifications were made to the equipment. Lung function testing apparatus designated for HIV-seropositive patients
was used throughout. One-way valve safety mouthpieces (Vitalograph Ltd., Buckingham) were used for spirometry to avoid the risk of inhaling pathogens from the spirometer. A Pall Ultipors breathing system filter (pall Biomedical, Havant, Portsmouth) was placed distal to the mouthpiece during the CO transfer test. The increased tidal volume of 150ml due to the presence of the filter was incorporated in the calculation. Plastic mouthpieces and nose clips were sterilized in 2% glutaraldehyde for 1 h between patients and the remainder of the apparatus dismantled and sterilized weekly for >3 h in 2% glutaraldehyde. For each patient, values of each lung function measurement were compared with those predicted for age, sex, and height (9). The results, expressed as a percentage of predicted for each patient, were grouped according to the clinical classification of patients, and statistical comparisons were made using the Mann-Whitney U test, the Wilcoxon matched-pairs signed-rank test, or analysis of variance (ANOVA) as indicated. Data are presented as mean ± standard error of the mean (SEM). The study was approved by he hospital ethics committee.
Results
Pulmonary hmction on Presentation A group of 4i4 patients werestudied. The number with each clinical diagnosis is shown in table 1. The age, percentage who were still smoking at the time of the study, and percentage who had mild or severe dyspnea in each patient group are listed in table 1. Nonsmokers were defined as those who had never smoked or those who had abstained from smoking for> 6 months. Mild dyspnea was record-
ed if the patient complained of breathlessness on stairs or hills, severe dyspnea if breathless on minimal exertion or at rest. Patients with acute PCP and lung mycobacterial disease had a high incidence of dyspnea. In the remaining groups up to 300,10 had mild dyspnea and approximately 10070 of all groups complained of severe dyspnea. Airway function (FEV 1 and PEF) was normal in most groups of HIV-infected patients. Patients with acute PCP had reduced values of FEV. and PEFR, and there were small reductions in patients with pulmonary KS (table 1). The mean result of pulmonary function tests on asymptomatic HIV-positive patients at the first visit revealed slight reductions in DLeo (880,10 of predicted) and Keo (92070 of predicted), but not FVC compared with predicted normal for the British population according to age, sex, height, and ethnic group (figure 1) (9). All other groups of patients were classified clinically at the time of the first lung function tests, and the results of the lung function measurements were compared with those from HIV-seropositiveasymptomatic patients. Patients with POL had a significant reduction in Keo (p < 0.01) and DLeo (820,10 of predicted: p < 0.05), and patients with ARC had a significant reduction in DLeo (730,10 of predicted), Keo, and FVC (p < 0.001, p < 0.001, and p < 0.001). Similarly, there were reductions in DLeo, Keo, and FVC in patients with nonpulmonary complications of AIDS. The DLeo (720,10 of predicted),
747
SERIAL PULMONARY FUNCTION IN HIV
100
T CO L
reductions in DLeo were also seen in the other categories of pulmonary disease complicating AIDS (pulmonary KS, 63070 of predicted, p< 0.001;mycobacterial infection, 68070 of predicted, p < 0.05; and pyogenic bacterial infection, 70070 of predicted, p < 0.05). The results of lung function tests on the presentation of these 474 patients include results on 169 patients that have been previously presented (3).
90 80 70 60 50 40
Keo 100 90 Q
W ~
o Ci W a: a.. 0~
80 70 60 50 40
120 110 100 90 80 70 60 50 40
HIV
PGL
ARC
+ve
n=
126
95
92
NON NON ACUTE POST LUNG LUNG LUNG PCP PCP KS KS AIDS 35
19
78
42
11
LUNG LUNG MYCO PYOG 11
7
Fig. 1. Lung function on presentation. OLeo. Keo, and FVC expressed as percentage of predicted in patients on entry into the study and classified according to clinical disease category (mean ± SEM). (***p < 0.001; **p < 0.Q1; *p < 0.05, Mann-Whitney U test.)
Kco, and FVC were reduced in AIDS patients with Kaposi's sarcoma but without pulmonary involvement (mainly cutaneous KS) (p < 0.001, p < 0.001, and p < 0.001)and in AIDS patients without KS and without pulmonary disease (opportunistic infections involving other organs), the DLeo (73070 of predicted), Keo,
and FVC were also reduced (p < 0.001, p < 0.001, and p < 0.05). In patients with AIDS and pulmonary complications, further reductions in lung function were 0 bserved. In patients with acute PCP marked reductions in mean Dteo, Keo, and FVC were found (p < 0.001, p < 0.001, and p < 0.001).Marked
Serial Lung Function over 18 Months in Clinically Stable Patients To dissociate subtle changes in clinically stable patients from changes associated with acute clinical deterioration, the results from serial lung function tests performed on the patients who did not change diagnostic group were analyzed separately. Pulmonary function tests were repeated at intervals of 3 months in patients from each category of HIV disease. Data were included from an 18month follow-up period or until the patient changed clinical category. In patients who did not change clinical diagnostic group, no significant reductions in DLeo, Keo, or FVC occurred in any patient group with the exception of pulmonary KS and post PCP (figure 2A). In patients with pulmonary KS there was some evidence of a gradual decline in FVC and DLeo. Following PCP there was a gradual improvement in DLeo (see later). Toensure that changes in individual patients were not obscured by grouping the data, the initial and 18-month values for individual patients who were followed throughout the study period and who did not change HIV category are presented separately (figure 2B). When compared by analysis of variance these individuals showed no tendency for a decline in the DLeo, with the exception of those individuals with lung KS. Lung Function in Patients with Clinical Deterioration Using the measurements from the 3month serial lung function tests, the results before and immediately following the deterioration were compared. Of the patients who deteriorated 41deteriorated such that they changed clinical diagnostic category during the period of the study. Of these patients, 8 who were initially asymptomatic HIV-seropositive patients or who had POL developed ARC. During this transition, there was no significant deterioration in DLeo or Keo; however, FVC results were more variable, with an increase in the four patients with POL who developed ARC
20
40
60
80
100
120
140
+~e I
6
9
12
15
18
23
16
14
91
I
8
15
7
15
18
101
18
17
12
I
18
120
20
40
o
3
6
MONTHS
20 I
40
60
15
4
60
12
5
80
~I
8
80
NON LUNG KS
In=35
100
~
140
,
3
o
43
i
3
o
25
6
ii,
39
6
6
6
,
o 3
6
i i i
7
3
i i
1n=11
20I
40
100
120
140
:11~
'
ln= 78
0
1N~N LUN~AI~I
3
140.-----In= 11 3 3 120
60
12
8
12
60
9
16
9
80
6
32
6
80
3
32
3
l
20 1
40
60
80
100
120
140
100
0
n= 87
0
IPGL
I
28
~
33
3
100
120
0
:1
60
9
6
4
15
i
12
5
12
,
3
15
3
15
,
3
18
12
I
1
15
i i i
7
,
3
I
12
2
i i '
14
9
i
3
I
120
140
20
40
60
80
100
120
140
80
~
20
40
~ 60
80
100
120
J
1
140
40
'#. 60
a..
~
~ 100
t> o
B
0
0
0
0
ARC
0
-
PGL
lr--
0
~
HIV+ve
-
18
18
I
--..,)01
....
18
Q
0
o
og....
LUNG KS
o
0-
~o
18
i
-0
0 . . . . - - - - - - -o
NON LUNG KS
40· 20
0
0
0
MONTHS
18
§=;:
