Brief Communications Prognosis of Idiopathic Pulmonary Fibrosis in Patients with Mucous Hypersecretion 1 • 2
N. HIWATARI, S. SHIMURA, T. SASAKI, T. AIKAWA, Y. ANDO, H. ISHIHARA, K. SEKIZAWA, H. SASAKI, and T. TAKISHIMA
Idiopathic pulmonary fibrosis (IPF) is characterized by recurrent or chronic inflammation in the alveolar regions that progresses to interstitial fibrosis. It has been stated that typical symptoms of IPF are exertional dyspnea with nonproductive cough and physiologic measurements indicating restrictive and diffusion defects. However, it is also true that some patients with IPF expectorate some amount of sputum even when no respiratory infection is found. In fact, Edwards and Carlile (1) have observed hyperplasia ofbronchial glands in a morphometric analysis of autopsied lungs from patients with IPF, and Turner-Warwick and coworkers (2)havefound that half of the patients with IPF from their study expectorated a significant amount of sputum. Our preliminary data from a morphometric analysis of the airways from autopsied lungs of patients with IPF also showed that more prominent inflammation and mucus plugging are observed in the airways of patients with mucous hypersecretion than in those without mucous hypersecretion (3). Although abnormalities in the airways in addition to the peripheral lung are supposed to be among the factors determining the survival rate in IPF, little attention has been paid to the airway lesions in IPF, and the significance of mucous hypersecretion as a clinical feature of patients with IPF is still unknown. In the present study, we observed that mucous hypersecretion in patients with chronic IPF suggests a poorer prognosis and also that mucous hypersecretion is accompanied by increased neutrophils and eosinophils in bronchoalveolar lavage fluid (BALF). Seventy patients were hospitalized at Tohoku University Hospital for further examination of suspected IPF between 1970 and 1988 because IPF had been the most likely diagnosis made on the basis of a prior visit to our clinic. The diagnosis of IPF was established in 50 patients. From 50 patients, 25 (nine women and 16 men with a mean age of 59 ± 3 yr) who survived beyond 1 yr and also had complete medical records of the subsequent examinations were chosen for the present study. The diagnosis of IPF was made from a combination of medical history, physical examination, laboratory tests, chest roentgenograms, pulmonary function tests, arterial blood gas analysis, and the results of lung biopsies according to previously described criteria (4,5). Detailed inpatient investigations, including skin tests, precipitating antibody
182
SUMMARY In order to determine the prognosis of patients with chronic idiopathic pulmonary fibrosis (IPF), we evaluated clinical, laboratory, and bronchoalveolar lavage (BAL) data at the onset of IPF in 25 patients who survived beyond 1 yr (nine women and 16 men, 59 ± 3 yr of age, mean ± SE). When the patients were divided into two groups according to whether they had or did not have mucous hypersecretion, 11 patients with hypersecretion (Group A) had a poorer survival rate (6 yr) than did 14 patients without hypersecretion (Group B) (10 yr) (p < 0.01). Further, there was a significant negative correlation between sputum volume and the duration of survival in 25 patients (r = - 0.55, P < 0.01).Before glucocorticoid treatment, we also found significantly larger numbers of neutrophils (17%)and eoslnophils (5%) in differential cell counts of bronchoalveolar lavage fluid (BALF) In Group A than in Group B (neutrophils, 1%; eoslnophils, 0.6%) (p < 0.05each). Chest radiographic findings and other laboratory data including pulmonary function tests did not correlate with the survival rate. These findings suggest that mucous hypersecretion as well as neutrophils and eosinophils in BALF are among the determinants of prognosis In patients with chronic IPF. AM REV RESPIR DIS 1991; 143:182-185
examination, and ophthalmologic and histologic examinations excluded sarcoidosis, hypersensitivity pneumonitis, pneumoconiosis, collagen diseases, and mixed connectivetissue disease. Histologic confirmation of the diagnosis was obtained by transbronchial lung biopsy (TBLB) in all patients and, in addition, two patients underwent open lung biopsy. No alteration in the diagnosis of IPF was found even after postmortem examination in any of the patients, except for the seven patients who were still alive in February 1989. The patients weredivided into two groups. Group A was made up of those with 10 mllday or more of sputum (two women and nine men with a mean age of 65 ± 3 yr). Group B was made up of those with no or little sputum (seven women and seven men with a mean age of 55 ± 4 yr). Seven of these 14patients (four women and three men with a mean age of 51 ± 5 yr) are still alive and have been followed for 8 yr or more in the Tohoku University Clinic as of February 1989. We took as the date of onset of IPF the time when the patients first noticed subjective symptoms such as exertional dyspnea, dry cough, sputum, and other general symptoms (including fever, general fatigue, and flulike symptoms) that wereconfirmed not to be from diseases other than IPF. Clinical, laboratory, and chest radiographic data used for the present study were all obtained at the first visit to the First Department of Internal Medicine Tohoku University School of Medicine. Pulmonary function tests, trans bronchial or open chest lung biopsy, and bronchoalveolar lavagewerealso performed during their first admission, prior to glucocorticoid therapy. The grade of dyspnea was estimated by the classification of Hugh-Jones (6). Sputum was collected in a scaled cup, and sputum volume per day was assessed as the mean of 6 days or more at those times when the patients were not suffering from
an exacerbation of their disease or from a pulmonary infection according to clinical and laboratory data. On the basis of their medical records, these sputum volumes were the smallest values for at least 1yr since the onset of disease.No significant pathogens were found in the sputum specimens during the period when the sputum volume was obtained for the present study. VC and FEV! were measured using a spirometer (OST80D; Chest Co., Tokyo, Japan). Pao2, Paco2, and pH were measured using a blood gas analyser (IL meter 213; Instrumentation Laboratories, Lexington, MA). Pulmonary diffusing capacity for carbon monoxide (DLco) was determined by the single-breath technique. VC and DLco were expressed as percentages of predicted values (7) and FEV! 010 was obtained by dividing FEV! by VC. Using a fiberoptic bronchoscope (BF 2TR; Olympus Co., Tokyo, Japan), four or more specimens from the peripheral regions of different unilateral lobes were obtained from each patient. The specimens wereembedded in paraffin after fixation with formalin, sectioned to 4 urnthick, and stained with hematoxylin-eosin, elastica-Goldner, and periodic
(Received in originalform December 27, 1989 and in revised form April 10, 1990) ! From the First Department of Internal Medicine, Tohoku University School of Medicine, Sendai, Japan. 2 Correspondence and requests for reprints should be addressed to Tamotsu Takishima, M.D., Professor and Chairman, First Department of Internal Medicine,Tohoku UniversitySchool of Medicine, 1-1 Seiryo-Machi, Aoba-ku, Sendai 980, Japan.
183
BRIEF COMMUNICATION
acid-Schiff. The TBLB samples for histologic examinations were about 3 x 3 mm in size and included mainly tissues from the alveolar region. After topical anesthesia, the tip of the bronchoscope was wedged into a segment or sub segment of the middle lobe or lingular bronchus. Sterile saline (0.90/0 at 37°C) was then instilled through the bronchoscope in 50-ml aliquots to a total of 150 ml, and the fluid was recovered immediately by gentle suctioning using low negative pressure. The fluids obtained were strained through one layer of surgical gauze and centrifuged for 8 min at 1,200 rpm. The cell pellets were used to count total numbers and smears were made to determine the differential cell count by means of Wright-Giemsa and nonspecific esterase stains. Cell counts were expressed as a percentage of the total cell count, excluding epithelial cells and erythrocytes. According to the criteria proposed by the project team for interstitial pulmonary disease organized in 1974and supported by the Japan Ministry of Welfare (8), we classified the chest radiographic findings into four categories: type I, nodular and/or patchy shadow (ground-glass appearance is also included in this type); type II, reticulonodular shadow; type III, honeycomb pattern; type IV, when the radiogram showed a decrease in lung volume. Data are presented as the mean ± SE. For mean comparisons, Student's two-tailed unpaired t test was used, and the Cochran-Cox t test was employed when Bartlett's test for uniformity of variance showed it to be nonuniform. Comparison between two groups on the basis of mucous hypersecretion was done using a table analysis and the MannWhitney rank test. For differential cell counts of BALF, Wilcoxon's rank test was used because of unbalanced deviation of samples (nonnormal deviation). The regression coefficient was also used for
statistical analysis; p < 0.05 was considered to be statistically significant.
The clinical and physicalcharacteristics of the patients with IPF in Groups A and B are summarized in table 1. There was no significant difference between women and men regarding the duration of survival (6 yr for women and 5 yr for men). There was a significant difference between Groups A and B regarding the duration of survival, as shown in table 1 (p < 0.01). Namely, those in Group A had a significantly shorter duration of survival (6 ± 0.6 yr) than did those in Group B (10 ± 0.9 yr). No significant difference in age at the onset of the disease was observed between Groups A and B (table 1). No significant difference between Groups A and B regarding the incidence of cigarette smoking was observed, as is shown in table 1. As shown in figure 1, there was a significant negative correlation between sputum volume and the duration of survival (r = - 0.55, p < 0.01). No significant differences between Groups A and B were observed in the laboratory data summarized in table 1. The incidence of positive C-reactive protein was higher in Group A than in Group B, although not significantly so. No significant difference was found between Groups Aand B in radiographic findings, as shown in table 1.
TABLE 1
(mil/day)
50
r == -0.55
40 Ql
E .2 30 o > E ::J
~ 20
Cf)
10
10
12
14
Survival
16 (yrs)
Fig. 1. Relationship between sputum volume (ml/day) and the duration of survival (years) in 25 patients of the present study. Sputum volume is significantly correlated with the duration of survival (r = -0.55, P < 0.01).
A significant increase in the neutrophil percentage of differential cellcounts from HALF was seen in Group A (17.± 9070) compared with that in Group B (1 ± 0.3070) (p < 0.05) (table 1 and figure 2, top panel). Similarly, a significant increase in eosinophil percentage in the HALF was seen in Group A (5 ± 2070) compared with that in Group B (0.6 ± 0.3070) (p < 0.05) (table 1 and figure 2, bottom panel). No significant differences in prognosis resulting from the treatments (including glucocorticoid) were observed between Groups
CLINICAL AND LABORATORY DATA IN GROUPS A AND B*
Sex Age at onset, yr Duration of survival, yr Grade of dyspnea t Dyspnea, % Cough, % Sputum volume, mllday Smokers, % Laboratory Sedimentation rate, mm/1 h LDH, lUlL C-reactive protein positive, % VC,% FEV1% Pao2, mm Hg DLco, % Radiographic findings* Types I and/or II, % Types III and/or IV, % Bronchoalveolar lavage Recovery of fluid, % Cell differentials Macrophage, % Lymphocyte, % Neutrophil, % Eosinophil, %
Group A (n = 11)
Group B (n = 14)
2F/9M 65 ± 3 5.8 ± 0.6§ 2.8 ± 0.3 100 100 24 ± 5§ 36 (1F/3M)
7F/7M 55 ± 4 9.7 ± 0.9 2.0 ± 0.2 79 86 0.7 ± 0.5 43 (2F/4M)
61 ± 11 401 ± 25 73 63 ± 5 84 ± 1 59 ± 4 47 ± 10
52 ± 11 399 ± 29 43 78 ± 9 84 ± 2 66 ± 3 62 ± 6
91 73
100 57
52 ± 11
~
..c Q.
0lo. +'"
:::s
Q)
z
± ± ± ±
12 2 911 211
91 7 1 0.6
± ± ± ±
4 4 0.3 0.3
30 25 20
*
15 10
5 0 ~
.c
Q.
0
C
en
0
w 72 6 17 5
* Data are shown as mean ± SE. of Hugh-Jones. :j: See text for the details of the classification of radiographic findings. § p < 0.01, significant difference between Groups A and B. II p < 0.05, significant difference between Groups A and B.
t Based on the classification
46 ± 5
(% )
Group A Group B
(0/0 )
1:~
.*
Group A Group B
Fig. 2. Percentages of neutrophils (upper panel) and eosinophils (lower panel) in differential cell counts from BALF in two groups of patients with IPF: Group A, patients with sputum of 10 ml or more per day; Group B, patients with no or little sputum. Group A showed significant increases in neutrophil and eosinophil percentages compared with Group B. Asterisks indicate p < 0.05.
184
BRIEF COMMUNICATION
A and B in the present study. For example, 10 of 11 patients (910/0) and 11 of 14 patients (77 %) had receivedglucocorticoid treatment in Groups A and B, respectively.
** * The pulmonary infiltrates wereinitially evaluated using the noninvasive tests described above.Thereafter, a TBLB was indicated, and an open lung biopsy was performed if the diagnosis of IPF wasnot established after histologic examination by TBLB. The diagnosis of IPF wasnot changed by postmortem histologic examination in all patients who died. This was also true of the other patients in whom IPF was diagnosed in our hospital. Similarly, some investigators (9) have established the diagnosis of IPF by TBLB and not necessarily by open lung biopsy. Therefore, it is unlikely that the diagnosis of IPF in the patients who are alive in the present study would be changed by open lung biopsy or postmortem examination. To date, there have been few studies concerning sputum production in IPF. In contrast to our observations, Turner-Warwick and coworkers (2) have reported that sputum production in patients with IPF has no significant relationship to survival, although sputum production was observed in about 60% of patients with IPF examined in their study. There are some differences between the selection of the patients of Turner-Warwick and coworkers (2) and ours. The main difference is that they included patients with an acute type of IPF with fulminant progress and features typical of Hamman-Rich syndrome for their study. For the present study, weexcluded patients with IPF who died within 1 yr from the onset of disease. The reason for this is that their fulminant progress made it difficult to distinguish the clinical and laboratory data from those of complications caused by conditions other than IPF itself. For example, persistent mucous hypersecretion was infrequently observedin patients with acute IPF because almost all of them died within a few months of the onset of disease. Even if mucous hypersecretion is estimated in these patients, it seems impossible to estimate its relationship to survival over a duration as short as 3 to 6 months. The other difference is that they included patients with IPF who were at different stages of the disease (early to terminal) for their study. It is possible that sputum appears at the terminal period in chronic IPF because of recurrent respiratory infection. We examined the effect of the presence of mucous hypersecretion on the survival in IPF with special attention to the presence of sputum in the early period when no infections are observed. To our knowledge, this is the first reported study of the significance ofmucous hypersecretion in the clinical features of IPF. Persistent mucous hypersecretion may increase the incidence of respiratory infection, airway inflammation, and obstructive impairment in patients with IPF as well as in those
suffering from other respiratory diseases, resulting in a shorter survival rate. Although the precise mechanisms responsible for increased mucus production in patients with IPF are not clear, we can address the following possibilities. First, prolonged smoking induces the mucous hypersecretion associated with enlargement of the mucussecreting tissue and mucosal inflammation of the airways (l0, 11). In the present study, however, the incidence of smoking did not differ significantly between the two groups, indicating that this possibility is unlikely. Second, it is possible that recurrent and persistent respiratory infection might produce hyperplasia of bronchial glands and mucous hypersecretion in patients with IPF. Regarding this point, Edwards and coworkers (1) reported that the quantity of gland and muscle was significantly greater in patients with IPF than in a group of normal control subjects in a morphometric study of the bronchi of autopsied lungs, and they speculated that these changes weredue to proximal extension from repeated and persistent infection of the lung parenchyma. Toexclude this possibility, we measured the sputum volume by including the data from the early period and also excluding the periods when pulmonary infection, even latent infection or common cold, was present. Third, it is known that various chemical mediators from mast cells and basophils produce an increase in mucous secretion from human airways (12-14). Kawanami and coworkers (15) have reported abundant mast cells in fibrotic lesions in the peripheral lung of patients with IPF, and increased histamine levelshave been measured in the BALF from patients with IPF (16). Therefore, these chemical mediators may in part play a role in mucous hypersecretion in IPF. Finally, various other inflammations in the airways are known to produce mucous hypersecretion in human lungs. The higher positive rate of C-reactive protein in Group A, although the difference was not statistically significant, may indicate an inflammation in the lung that reflects the genesis of IPF since clinical and other laboratory data excluded possible infections. Although we have no direct evidence,it is possible that these inflammations in the peripheral regions extend to the proximal airways, resulting in mucous hypersecretion.However, there havebeen no reports concerning the effects of mucous hypersecretion on histologic aspects of autopsied lungs from patients with IPF. The reason for this may be that clinical features other than mucous hypersecretion before autopsy affect the histologic aspects, concealing the effect of mucous hypersecretion. Therefore, we attempted to look for the effect of mucous hypersecretion on the morphologic features of airways of patients with IPF by excluding the possibility of effects other than mucous hypersecretion if possible, in autopsied lungs of patients with IPF different from those of the present study. We compared the histologic aspects of autopsied lungs from patients with IPF and mucous hypersecretion
for 3 yr or more before death with that from patients without mucous hypersecretion for at least 3 yr, matching the clinical features other than mucous hypersecretion. We found severer inflammation and mucus plugging in airways from patients with mucous hypersecretionthan in those without mucous hypersecretion in our preliminary study (3). Haslam and coworkers (17), Rudd and colleagues (18), and Turner-Warwick and coworkers (19) have demonstrated that an increase of eosinophils or neutrophils or both with a smaller lymphocyte count in HALF is related to a poor clinical response to corticosteroids. Watters and coworkers (20) also confirmed these results, and they demonstrated that the HALF eosinophil content correlated significantly with the severity of clinical impairment. Further, Hallgren and coworkers (21) suggested a common activator of eosinophils and neutrophils in the development of lung tissue damage. In our present study, cellular analysis of BALF showed increases in neutrophils and eosinophils in Group A, and this is consistent with the observations described above. However, no attention has been given to the relationship between BALF findings and mucous hypersecretion in patients with IPF until the present study. Although the relationship between mucous hypersecretion and inflammatory cells in BALF is not directly discussed and confirmed, there is the possibility that various chemical mediators derived from inflammatory cells promote mucous hypersecretion in part, resulting in a shorter survival rate.
Acknowledgment The writers thank Dr. Ronald Scott for reading and Miss Kumiko Shibuya for preparation of the manuscript.
References 1. Edwards CW, Carlile A. The larger bronchi in cryptogenic fibrosing alveolitis: a morphometric study. Thorax 1982; 37:828-33. 2. Turner-Warwick M, Burrows B, Jonson A. Cryptogenic fibrosing alveolitis: clinical features and their influence on survival. Thorax 1980; 35: 171-80. 3. Ando Y, Aikawa T, Shimura S, Sasaki H, Takishima T. Morphometry of bronchial glands, goblet cells and intraluminal mucus in the patients with idiopathic pulmonary fibrosis with hypersecretion (abstract). Am RevRespir Dis 1989;139:A293. 4. Crystal RO, Fulmer JD, Roberts WC, Moss ML, Line BR, Reynolds HY. Idiopathic pulmonary fibrosis: clinical, histologic, radiographic, physiologic, scintigraphic, cytologic and biochemical aspects. Ann Intern Med 1976; 85:769-88. 5. Winterbauer RH, Hammer SP, Hallman KO, et al. Diffuse interstitial pneumonitis. Clinicopathologic correlations in 20 patients treated with prednisone/azathioprine. Am J Med 1978; 5:661-72. 6. Hugh-Jones P. A simple standard exercise test and its use for measuring exertion dyspnoea. Br Med J 1952; 1:65-71. 7. Cotes lE. Lung function. 4th ed. London: Blackwell, 1979; 369. 8. Nobeji T. Radiographic findings and radiographic examination. In: Murao M, ed. Idiopathic pulmonary fibrosis and related disorders. Research
185
BRIEF COMMUNICATION
report of Ministry of Welfare: Report from research group of pulmonary fibrosis. Tokyo: Ministry of Welfare, 1980; 21-35. 9. Karpel JP, Norin AJ. Association of activated cytolytic lung lymphocytes with response to prednisone therapy in patients with idiopathic pulmonary fibrosis. Chest 1989; 96:794-8. 10. Mullen JBM, Wright JL, Wiggs BR, Pare PD, Hogg JC. Structure of central airways in current smokers and exsmokers with and without mucus hypersecretion: relationship to lung function. Thorax 1987; 42:843-8. 11. Cosio MG, Hale KA, Niewoehner DE. Morphologic and morphometric effects of prolonged cigarette smoking on the small airways. Am Rev Respir Dis 1980; 122:265-71. 12. Shelhamer JH, Marom Z, Kaliner M. Immunologic and neuropharmacologic stimulation of mucous glycoprotein release from human airways in vitro. J Clin Invest 1980; 66:1400-8.
13. Marom Z, Shelhamer JH, Bach MK, Morton DR, Kaliner M. Slow-reacting substances, leukotrienes C 4 and D 4 , increase the release of mucus from human airways in vitro. Am Rev Respir Dis 1982; 126:449-51. 14. Marom Z, Shelhamer JH, Sun F, Kaliner M. Human airway monohydroxy-eicosatetraenoic acid generation and mucus release. J Clin Invest 1983; 72:122-7. 15. Kawanami 0, Ferrans VJ, Fulmer JD, Crystal RG. Ultrastructure of pulmonary mast cells in patients with fibrotic lung disorders. Lab Invest 1979; 40:717-34. 16. Casale TB, Trapp S, Zehr B, Hunninghake GW. Bronchoalveolar lavage fluid histamine levels in interstitiallung diseases. Am Rev Respir Dis 1988; 138:1604-8. 17. Haslam PL, Turton CWG, Lukoszek A, et al. Bronchoalveolar lavage fluid cell counts in cryptogenic fibrosing alveolitis and their relation to ther-
apy. Thorax 1980; 35:328-39. 18. Rudd RM, Haslam PL, Turner-Warwick M. Cryptogenic fibrosing alveolitis: relationships of pulmonary physiology and bronchoalveolar lavage to response to treatment and prognosis. Am Rev Respir Dis 1981; 124:1-8. 19. Turner-Warwick M, Burrows B, Johnson A. Cryptogenic fibrosing alveolitis: response to corticosteroid treatment and its effect on survival. Thorax 1980; 35:593-9. 20. Watters LC, Schwarz MI, Cherniack RM, et al. Idiopathic pulmonary fibrosis: pretreatment bronchoalveolar lavage cellular constituents and their relationships with lung histopathology and clinical response to therapy. Am Rev Respir Dis 1987; 135:696-704. 21. Hallgren R, Bjermer L, Lundgren R, Venge P. The eosinophil component of the alveolitis in idiopathic pulmonary hibrosis. Am Rev Respir Dis 1989; 139:373-7.
The Influence of Human Immunodeficiency Virus Infection on Tuberculosis in Kampala, Uganda 1 - 3
PETER R ERIKI, ALPHONSE OKWERA, TOM AISU, ANNE B. MORRISSEY, JERROLD J. ELLNER, and THOMAS M. DANIEL
Tuberculosis is recognized as a major infectious complication of the acquired immunodeficiency syndrome (AIDS) (1, 2). Clinical studies have emphasized the early appearance of tuberculosis in the course of AIDS, the frequency of extrapulmonary tuberculosis in patients with AIDS, and the occurrence of nontypical radiographic patterns in pulmonary tuberculosis in patients with AIDS. In Africa, tuberculosis represents the most important pulmonary complication seen in patients with AIDS (3, 4). The prevalence of HIV infection in Uganda, East Africa is among the highest in the world (5-7). Recent Ministry of Health estimates (6) place the general population HIVseropositivity rate for the country at 6070 and for Kampala, the capital city, at 17070. Seropositivity rates exceed 50070 in some young adult groups (7), the age cohort in which most tuberculosis occurs. There are no current estimates of tuberculosis incidence or prevalence in Uganda. Thberculin surveys in 1958 and 1970 found approximately 65070 of adults tuberculin positive with no change between the two surveys (8). In this report, we cite clinical, radiographic, and microbiologic fmdings on 59 patients with pulmonary tuberculosis seen at Old Mulago Hospital, Kampala, Uganda, two-thirds of whom were found to be HIV-seropositive. Fifty-nine adult patients with pulmonary tuberculosis consecutively admitted to Old Mulago Hospital between November 1988 and May 1989 were studied. All patients admitted to the hospital were
SUMMARY The clinical, radiographic, and microbiologic features of 59 patients with pulmonary tuberculosis In Kampala, Uganda were studied and correlated with the serologic reactivity to the human Immunodeficiency virus (HIV) of these patients. 'TWo-thirds of the patients with tuberculosis were HIY-seroposltlve. Histories of fever and weight loss were more prominent In HIY-seroposltlve patients, and perlhllar and basilar Infiltrative diseases were more frequently seen In HIY-seroposltlve patients. Although all patients responded similarly to drug therapy, cutaneous drug reactions were seen In nearly one-third of HIY-seroposltlve patients receiving thiacetazone. AM REV RESPIR DIS 1991; 143:185-187
included in the study if acid-fast bacilli were seen on smear or Mycobacterium tuberculosis was recovered in the culture of at least one sputum sample. One patient with pulmonary disease consistent with active tuberculosis in whom the diagnosis was confirmed by histologic examination of a scrofulous lymph node was also included. Clinical assessment was based upon history and physical examination by one of the investigators (AO). All patients except for one woman in the first trimester of pregnancy had posteroanterior chest radiographs that were reviewed by two of the investigators (PPE and AO). Radiographs were classified as typical if the disease was predominantly an upper lobe disease of a fibronodular, exudative, or cavitary nature and as not typical if the disease was predominantly perihilar, including hilar adenopathy, or basilar and of an infiltrative nature. Expectorated sputum specimens were examined at the National Tuberculosis Reference Laboratory at Wandageya. Sputa were digested with 4010 sodium hydroxide, followed by neutralization with hydrochloric acid. After centrifugation, a sample of the pellet was stained with auramine 0 and examined by fluorescence microscopy. The remainder of the pellet was cultured on a Lowenstein-Jensen
medium. All isolates were tested for niacin production using niacin test strips (Difco Laboratories, Detroit, MI). HIV serologic status was assessed by an ELISA of serum antibody using Wellcozyme anti-HTLV III kits (Wellcome Diagnostics, Dartford, England) or recombigen-HIV EIA kits (Cambridge BioScience, Worcester, MA). All positive HIV ELISA serologies and a 10% random sample of negative serologies were confirmed by western immunoblotting using the novapath immunoblot assay (Biorad Laboratories, Hercules, CAl.
(Received in original form June 8, 1990 and in revised form August 20, 1990) 1 From the Ministry of Health, Thberculosis Control Programme, and Makerere University, Kampala, Uganda and from Case Western Reserve University, Cleveland, Ohio. 2 Supported in part by Grant No. AI -26482 from the National Institute of Allergy and Infectious Diseases, National Institutes of Health. 3 Correspondence and requests for reprints should be addressed to Thomas M. Daniel, M.D., Department of Medicine, University Hospitals, 2074 Abington Road, Cleveland, OH 22106.
N. HIWATARI, S. SHIMURA, T. SASAKI, T. AIKAWA, Y. ANDO, H. ISHIHARA, K. SEKIZAWA, H. SASAKI, and T. TAKISHIMA
Idiopathic pulmonary fibrosis (IPF) is characterized by recurrent or chronic inflammation in the alveolar regions that progresses to interstitial fibrosis. It has been stated that typical symptoms of IPF are exertional dyspnea with nonproductive cough and physiologic measurements indicating restrictive and diffusion defects. However, it is also true that some patients with IPF expectorate some amount of sputum even when no respiratory infection is found. In fact, Edwards and Carlile (1) have observed hyperplasia ofbronchial glands in a morphometric analysis of autopsied lungs from patients with IPF, and Turner-Warwick and coworkers (2)havefound that half of the patients with IPF from their study expectorated a significant amount of sputum. Our preliminary data from a morphometric analysis of the airways from autopsied lungs of patients with IPF also showed that more prominent inflammation and mucus plugging are observed in the airways of patients with mucous hypersecretion than in those without mucous hypersecretion (3). Although abnormalities in the airways in addition to the peripheral lung are supposed to be among the factors determining the survival rate in IPF, little attention has been paid to the airway lesions in IPF, and the significance of mucous hypersecretion as a clinical feature of patients with IPF is still unknown. In the present study, we observed that mucous hypersecretion in patients with chronic IPF suggests a poorer prognosis and also that mucous hypersecretion is accompanied by increased neutrophils and eosinophils in bronchoalveolar lavage fluid (BALF). Seventy patients were hospitalized at Tohoku University Hospital for further examination of suspected IPF between 1970 and 1988 because IPF had been the most likely diagnosis made on the basis of a prior visit to our clinic. The diagnosis of IPF was established in 50 patients. From 50 patients, 25 (nine women and 16 men with a mean age of 59 ± 3 yr) who survived beyond 1 yr and also had complete medical records of the subsequent examinations were chosen for the present study. The diagnosis of IPF was made from a combination of medical history, physical examination, laboratory tests, chest roentgenograms, pulmonary function tests, arterial blood gas analysis, and the results of lung biopsies according to previously described criteria (4,5). Detailed inpatient investigations, including skin tests, precipitating antibody
182
SUMMARY In order to determine the prognosis of patients with chronic idiopathic pulmonary fibrosis (IPF), we evaluated clinical, laboratory, and bronchoalveolar lavage (BAL) data at the onset of IPF in 25 patients who survived beyond 1 yr (nine women and 16 men, 59 ± 3 yr of age, mean ± SE). When the patients were divided into two groups according to whether they had or did not have mucous hypersecretion, 11 patients with hypersecretion (Group A) had a poorer survival rate (6 yr) than did 14 patients without hypersecretion (Group B) (10 yr) (p < 0.01). Further, there was a significant negative correlation between sputum volume and the duration of survival in 25 patients (r = - 0.55, P < 0.01).Before glucocorticoid treatment, we also found significantly larger numbers of neutrophils (17%)and eoslnophils (5%) in differential cell counts of bronchoalveolar lavage fluid (BALF) In Group A than in Group B (neutrophils, 1%; eoslnophils, 0.6%) (p < 0.05each). Chest radiographic findings and other laboratory data including pulmonary function tests did not correlate with the survival rate. These findings suggest that mucous hypersecretion as well as neutrophils and eosinophils in BALF are among the determinants of prognosis In patients with chronic IPF. AM REV RESPIR DIS 1991; 143:182-185
examination, and ophthalmologic and histologic examinations excluded sarcoidosis, hypersensitivity pneumonitis, pneumoconiosis, collagen diseases, and mixed connectivetissue disease. Histologic confirmation of the diagnosis was obtained by transbronchial lung biopsy (TBLB) in all patients and, in addition, two patients underwent open lung biopsy. No alteration in the diagnosis of IPF was found even after postmortem examination in any of the patients, except for the seven patients who were still alive in February 1989. The patients weredivided into two groups. Group A was made up of those with 10 mllday or more of sputum (two women and nine men with a mean age of 65 ± 3 yr). Group B was made up of those with no or little sputum (seven women and seven men with a mean age of 55 ± 4 yr). Seven of these 14patients (four women and three men with a mean age of 51 ± 5 yr) are still alive and have been followed for 8 yr or more in the Tohoku University Clinic as of February 1989. We took as the date of onset of IPF the time when the patients first noticed subjective symptoms such as exertional dyspnea, dry cough, sputum, and other general symptoms (including fever, general fatigue, and flulike symptoms) that wereconfirmed not to be from diseases other than IPF. Clinical, laboratory, and chest radiographic data used for the present study were all obtained at the first visit to the First Department of Internal Medicine Tohoku University School of Medicine. Pulmonary function tests, trans bronchial or open chest lung biopsy, and bronchoalveolar lavagewerealso performed during their first admission, prior to glucocorticoid therapy. The grade of dyspnea was estimated by the classification of Hugh-Jones (6). Sputum was collected in a scaled cup, and sputum volume per day was assessed as the mean of 6 days or more at those times when the patients were not suffering from
an exacerbation of their disease or from a pulmonary infection according to clinical and laboratory data. On the basis of their medical records, these sputum volumes were the smallest values for at least 1yr since the onset of disease.No significant pathogens were found in the sputum specimens during the period when the sputum volume was obtained for the present study. VC and FEV! were measured using a spirometer (OST80D; Chest Co., Tokyo, Japan). Pao2, Paco2, and pH were measured using a blood gas analyser (IL meter 213; Instrumentation Laboratories, Lexington, MA). Pulmonary diffusing capacity for carbon monoxide (DLco) was determined by the single-breath technique. VC and DLco were expressed as percentages of predicted values (7) and FEV! 010 was obtained by dividing FEV! by VC. Using a fiberoptic bronchoscope (BF 2TR; Olympus Co., Tokyo, Japan), four or more specimens from the peripheral regions of different unilateral lobes were obtained from each patient. The specimens wereembedded in paraffin after fixation with formalin, sectioned to 4 urnthick, and stained with hematoxylin-eosin, elastica-Goldner, and periodic
(Received in originalform December 27, 1989 and in revised form April 10, 1990) ! From the First Department of Internal Medicine, Tohoku University School of Medicine, Sendai, Japan. 2 Correspondence and requests for reprints should be addressed to Tamotsu Takishima, M.D., Professor and Chairman, First Department of Internal Medicine,Tohoku UniversitySchool of Medicine, 1-1 Seiryo-Machi, Aoba-ku, Sendai 980, Japan.
183
BRIEF COMMUNICATION
acid-Schiff. The TBLB samples for histologic examinations were about 3 x 3 mm in size and included mainly tissues from the alveolar region. After topical anesthesia, the tip of the bronchoscope was wedged into a segment or sub segment of the middle lobe or lingular bronchus. Sterile saline (0.90/0 at 37°C) was then instilled through the bronchoscope in 50-ml aliquots to a total of 150 ml, and the fluid was recovered immediately by gentle suctioning using low negative pressure. The fluids obtained were strained through one layer of surgical gauze and centrifuged for 8 min at 1,200 rpm. The cell pellets were used to count total numbers and smears were made to determine the differential cell count by means of Wright-Giemsa and nonspecific esterase stains. Cell counts were expressed as a percentage of the total cell count, excluding epithelial cells and erythrocytes. According to the criteria proposed by the project team for interstitial pulmonary disease organized in 1974and supported by the Japan Ministry of Welfare (8), we classified the chest radiographic findings into four categories: type I, nodular and/or patchy shadow (ground-glass appearance is also included in this type); type II, reticulonodular shadow; type III, honeycomb pattern; type IV, when the radiogram showed a decrease in lung volume. Data are presented as the mean ± SE. For mean comparisons, Student's two-tailed unpaired t test was used, and the Cochran-Cox t test was employed when Bartlett's test for uniformity of variance showed it to be nonuniform. Comparison between two groups on the basis of mucous hypersecretion was done using a table analysis and the MannWhitney rank test. For differential cell counts of BALF, Wilcoxon's rank test was used because of unbalanced deviation of samples (nonnormal deviation). The regression coefficient was also used for
statistical analysis; p < 0.05 was considered to be statistically significant.
The clinical and physicalcharacteristics of the patients with IPF in Groups A and B are summarized in table 1. There was no significant difference between women and men regarding the duration of survival (6 yr for women and 5 yr for men). There was a significant difference between Groups A and B regarding the duration of survival, as shown in table 1 (p < 0.01). Namely, those in Group A had a significantly shorter duration of survival (6 ± 0.6 yr) than did those in Group B (10 ± 0.9 yr). No significant difference in age at the onset of the disease was observed between Groups A and B (table 1). No significant difference between Groups A and B regarding the incidence of cigarette smoking was observed, as is shown in table 1. As shown in figure 1, there was a significant negative correlation between sputum volume and the duration of survival (r = - 0.55, p < 0.01). No significant differences between Groups A and B were observed in the laboratory data summarized in table 1. The incidence of positive C-reactive protein was higher in Group A than in Group B, although not significantly so. No significant difference was found between Groups Aand B in radiographic findings, as shown in table 1.
TABLE 1
(mil/day)
50
r == -0.55
40 Ql
E .2 30 o > E ::J
~ 20
Cf)
10
10
12
14
Survival
16 (yrs)
Fig. 1. Relationship between sputum volume (ml/day) and the duration of survival (years) in 25 patients of the present study. Sputum volume is significantly correlated with the duration of survival (r = -0.55, P < 0.01).
A significant increase in the neutrophil percentage of differential cellcounts from HALF was seen in Group A (17.± 9070) compared with that in Group B (1 ± 0.3070) (p < 0.05) (table 1 and figure 2, top panel). Similarly, a significant increase in eosinophil percentage in the HALF was seen in Group A (5 ± 2070) compared with that in Group B (0.6 ± 0.3070) (p < 0.05) (table 1 and figure 2, bottom panel). No significant differences in prognosis resulting from the treatments (including glucocorticoid) were observed between Groups
CLINICAL AND LABORATORY DATA IN GROUPS A AND B*
Sex Age at onset, yr Duration of survival, yr Grade of dyspnea t Dyspnea, % Cough, % Sputum volume, mllday Smokers, % Laboratory Sedimentation rate, mm/1 h LDH, lUlL C-reactive protein positive, % VC,% FEV1% Pao2, mm Hg DLco, % Radiographic findings* Types I and/or II, % Types III and/or IV, % Bronchoalveolar lavage Recovery of fluid, % Cell differentials Macrophage, % Lymphocyte, % Neutrophil, % Eosinophil, %
Group A (n = 11)
Group B (n = 14)
2F/9M 65 ± 3 5.8 ± 0.6§ 2.8 ± 0.3 100 100 24 ± 5§ 36 (1F/3M)
7F/7M 55 ± 4 9.7 ± 0.9 2.0 ± 0.2 79 86 0.7 ± 0.5 43 (2F/4M)
61 ± 11 401 ± 25 73 63 ± 5 84 ± 1 59 ± 4 47 ± 10
52 ± 11 399 ± 29 43 78 ± 9 84 ± 2 66 ± 3 62 ± 6
91 73
100 57
52 ± 11
~
..c Q.
0lo. +'"
:::s
Q)
z
± ± ± ±
12 2 911 211
91 7 1 0.6
± ± ± ±
4 4 0.3 0.3
30 25 20
*
15 10
5 0 ~
.c
Q.
0
C
en
0
w 72 6 17 5
* Data are shown as mean ± SE. of Hugh-Jones. :j: See text for the details of the classification of radiographic findings. § p < 0.01, significant difference between Groups A and B. II p < 0.05, significant difference between Groups A and B.
t Based on the classification
46 ± 5
(% )
Group A Group B
(0/0 )
1:~
.*
Group A Group B
Fig. 2. Percentages of neutrophils (upper panel) and eosinophils (lower panel) in differential cell counts from BALF in two groups of patients with IPF: Group A, patients with sputum of 10 ml or more per day; Group B, patients with no or little sputum. Group A showed significant increases in neutrophil and eosinophil percentages compared with Group B. Asterisks indicate p < 0.05.
184
BRIEF COMMUNICATION
A and B in the present study. For example, 10 of 11 patients (910/0) and 11 of 14 patients (77 %) had receivedglucocorticoid treatment in Groups A and B, respectively.
** * The pulmonary infiltrates wereinitially evaluated using the noninvasive tests described above.Thereafter, a TBLB was indicated, and an open lung biopsy was performed if the diagnosis of IPF wasnot established after histologic examination by TBLB. The diagnosis of IPF wasnot changed by postmortem histologic examination in all patients who died. This was also true of the other patients in whom IPF was diagnosed in our hospital. Similarly, some investigators (9) have established the diagnosis of IPF by TBLB and not necessarily by open lung biopsy. Therefore, it is unlikely that the diagnosis of IPF in the patients who are alive in the present study would be changed by open lung biopsy or postmortem examination. To date, there have been few studies concerning sputum production in IPF. In contrast to our observations, Turner-Warwick and coworkers (2) have reported that sputum production in patients with IPF has no significant relationship to survival, although sputum production was observed in about 60% of patients with IPF examined in their study. There are some differences between the selection of the patients of Turner-Warwick and coworkers (2) and ours. The main difference is that they included patients with an acute type of IPF with fulminant progress and features typical of Hamman-Rich syndrome for their study. For the present study, weexcluded patients with IPF who died within 1 yr from the onset of disease. The reason for this is that their fulminant progress made it difficult to distinguish the clinical and laboratory data from those of complications caused by conditions other than IPF itself. For example, persistent mucous hypersecretion was infrequently observedin patients with acute IPF because almost all of them died within a few months of the onset of disease. Even if mucous hypersecretion is estimated in these patients, it seems impossible to estimate its relationship to survival over a duration as short as 3 to 6 months. The other difference is that they included patients with IPF who were at different stages of the disease (early to terminal) for their study. It is possible that sputum appears at the terminal period in chronic IPF because of recurrent respiratory infection. We examined the effect of the presence of mucous hypersecretion on the survival in IPF with special attention to the presence of sputum in the early period when no infections are observed. To our knowledge, this is the first reported study of the significance ofmucous hypersecretion in the clinical features of IPF. Persistent mucous hypersecretion may increase the incidence of respiratory infection, airway inflammation, and obstructive impairment in patients with IPF as well as in those
suffering from other respiratory diseases, resulting in a shorter survival rate. Although the precise mechanisms responsible for increased mucus production in patients with IPF are not clear, we can address the following possibilities. First, prolonged smoking induces the mucous hypersecretion associated with enlargement of the mucussecreting tissue and mucosal inflammation of the airways (l0, 11). In the present study, however, the incidence of smoking did not differ significantly between the two groups, indicating that this possibility is unlikely. Second, it is possible that recurrent and persistent respiratory infection might produce hyperplasia of bronchial glands and mucous hypersecretion in patients with IPF. Regarding this point, Edwards and coworkers (1) reported that the quantity of gland and muscle was significantly greater in patients with IPF than in a group of normal control subjects in a morphometric study of the bronchi of autopsied lungs, and they speculated that these changes weredue to proximal extension from repeated and persistent infection of the lung parenchyma. Toexclude this possibility, we measured the sputum volume by including the data from the early period and also excluding the periods when pulmonary infection, even latent infection or common cold, was present. Third, it is known that various chemical mediators from mast cells and basophils produce an increase in mucous secretion from human airways (12-14). Kawanami and coworkers (15) have reported abundant mast cells in fibrotic lesions in the peripheral lung of patients with IPF, and increased histamine levelshave been measured in the BALF from patients with IPF (16). Therefore, these chemical mediators may in part play a role in mucous hypersecretion in IPF. Finally, various other inflammations in the airways are known to produce mucous hypersecretion in human lungs. The higher positive rate of C-reactive protein in Group A, although the difference was not statistically significant, may indicate an inflammation in the lung that reflects the genesis of IPF since clinical and other laboratory data excluded possible infections. Although we have no direct evidence,it is possible that these inflammations in the peripheral regions extend to the proximal airways, resulting in mucous hypersecretion.However, there havebeen no reports concerning the effects of mucous hypersecretion on histologic aspects of autopsied lungs from patients with IPF. The reason for this may be that clinical features other than mucous hypersecretion before autopsy affect the histologic aspects, concealing the effect of mucous hypersecretion. Therefore, we attempted to look for the effect of mucous hypersecretion on the morphologic features of airways of patients with IPF by excluding the possibility of effects other than mucous hypersecretion if possible, in autopsied lungs of patients with IPF different from those of the present study. We compared the histologic aspects of autopsied lungs from patients with IPF and mucous hypersecretion
for 3 yr or more before death with that from patients without mucous hypersecretion for at least 3 yr, matching the clinical features other than mucous hypersecretion. We found severer inflammation and mucus plugging in airways from patients with mucous hypersecretionthan in those without mucous hypersecretion in our preliminary study (3). Haslam and coworkers (17), Rudd and colleagues (18), and Turner-Warwick and coworkers (19) have demonstrated that an increase of eosinophils or neutrophils or both with a smaller lymphocyte count in HALF is related to a poor clinical response to corticosteroids. Watters and coworkers (20) also confirmed these results, and they demonstrated that the HALF eosinophil content correlated significantly with the severity of clinical impairment. Further, Hallgren and coworkers (21) suggested a common activator of eosinophils and neutrophils in the development of lung tissue damage. In our present study, cellular analysis of BALF showed increases in neutrophils and eosinophils in Group A, and this is consistent with the observations described above. However, no attention has been given to the relationship between BALF findings and mucous hypersecretion in patients with IPF until the present study. Although the relationship between mucous hypersecretion and inflammatory cells in BALF is not directly discussed and confirmed, there is the possibility that various chemical mediators derived from inflammatory cells promote mucous hypersecretion in part, resulting in a shorter survival rate.
Acknowledgment The writers thank Dr. Ronald Scott for reading and Miss Kumiko Shibuya for preparation of the manuscript.
References 1. Edwards CW, Carlile A. The larger bronchi in cryptogenic fibrosing alveolitis: a morphometric study. Thorax 1982; 37:828-33. 2. Turner-Warwick M, Burrows B, Jonson A. Cryptogenic fibrosing alveolitis: clinical features and their influence on survival. Thorax 1980; 35: 171-80. 3. Ando Y, Aikawa T, Shimura S, Sasaki H, Takishima T. Morphometry of bronchial glands, goblet cells and intraluminal mucus in the patients with idiopathic pulmonary fibrosis with hypersecretion (abstract). Am RevRespir Dis 1989;139:A293. 4. Crystal RO, Fulmer JD, Roberts WC, Moss ML, Line BR, Reynolds HY. Idiopathic pulmonary fibrosis: clinical, histologic, radiographic, physiologic, scintigraphic, cytologic and biochemical aspects. Ann Intern Med 1976; 85:769-88. 5. Winterbauer RH, Hammer SP, Hallman KO, et al. Diffuse interstitial pneumonitis. Clinicopathologic correlations in 20 patients treated with prednisone/azathioprine. Am J Med 1978; 5:661-72. 6. Hugh-Jones P. A simple standard exercise test and its use for measuring exertion dyspnoea. Br Med J 1952; 1:65-71. 7. Cotes lE. Lung function. 4th ed. London: Blackwell, 1979; 369. 8. Nobeji T. Radiographic findings and radiographic examination. In: Murao M, ed. Idiopathic pulmonary fibrosis and related disorders. Research
185
BRIEF COMMUNICATION
report of Ministry of Welfare: Report from research group of pulmonary fibrosis. Tokyo: Ministry of Welfare, 1980; 21-35. 9. Karpel JP, Norin AJ. Association of activated cytolytic lung lymphocytes with response to prednisone therapy in patients with idiopathic pulmonary fibrosis. Chest 1989; 96:794-8. 10. Mullen JBM, Wright JL, Wiggs BR, Pare PD, Hogg JC. Structure of central airways in current smokers and exsmokers with and without mucus hypersecretion: relationship to lung function. Thorax 1987; 42:843-8. 11. Cosio MG, Hale KA, Niewoehner DE. Morphologic and morphometric effects of prolonged cigarette smoking on the small airways. Am Rev Respir Dis 1980; 122:265-71. 12. Shelhamer JH, Marom Z, Kaliner M. Immunologic and neuropharmacologic stimulation of mucous glycoprotein release from human airways in vitro. J Clin Invest 1980; 66:1400-8.
13. Marom Z, Shelhamer JH, Bach MK, Morton DR, Kaliner M. Slow-reacting substances, leukotrienes C 4 and D 4 , increase the release of mucus from human airways in vitro. Am Rev Respir Dis 1982; 126:449-51. 14. Marom Z, Shelhamer JH, Sun F, Kaliner M. Human airway monohydroxy-eicosatetraenoic acid generation and mucus release. J Clin Invest 1983; 72:122-7. 15. Kawanami 0, Ferrans VJ, Fulmer JD, Crystal RG. Ultrastructure of pulmonary mast cells in patients with fibrotic lung disorders. Lab Invest 1979; 40:717-34. 16. Casale TB, Trapp S, Zehr B, Hunninghake GW. Bronchoalveolar lavage fluid histamine levels in interstitiallung diseases. Am Rev Respir Dis 1988; 138:1604-8. 17. Haslam PL, Turton CWG, Lukoszek A, et al. Bronchoalveolar lavage fluid cell counts in cryptogenic fibrosing alveolitis and their relation to ther-
apy. Thorax 1980; 35:328-39. 18. Rudd RM, Haslam PL, Turner-Warwick M. Cryptogenic fibrosing alveolitis: relationships of pulmonary physiology and bronchoalveolar lavage to response to treatment and prognosis. Am Rev Respir Dis 1981; 124:1-8. 19. Turner-Warwick M, Burrows B, Johnson A. Cryptogenic fibrosing alveolitis: response to corticosteroid treatment and its effect on survival. Thorax 1980; 35:593-9. 20. Watters LC, Schwarz MI, Cherniack RM, et al. Idiopathic pulmonary fibrosis: pretreatment bronchoalveolar lavage cellular constituents and their relationships with lung histopathology and clinical response to therapy. Am Rev Respir Dis 1987; 135:696-704. 21. Hallgren R, Bjermer L, Lundgren R, Venge P. The eosinophil component of the alveolitis in idiopathic pulmonary hibrosis. Am Rev Respir Dis 1989; 139:373-7.
The Influence of Human Immunodeficiency Virus Infection on Tuberculosis in Kampala, Uganda 1 - 3
PETER R ERIKI, ALPHONSE OKWERA, TOM AISU, ANNE B. MORRISSEY, JERROLD J. ELLNER, and THOMAS M. DANIEL
Tuberculosis is recognized as a major infectious complication of the acquired immunodeficiency syndrome (AIDS) (1, 2). Clinical studies have emphasized the early appearance of tuberculosis in the course of AIDS, the frequency of extrapulmonary tuberculosis in patients with AIDS, and the occurrence of nontypical radiographic patterns in pulmonary tuberculosis in patients with AIDS. In Africa, tuberculosis represents the most important pulmonary complication seen in patients with AIDS (3, 4). The prevalence of HIV infection in Uganda, East Africa is among the highest in the world (5-7). Recent Ministry of Health estimates (6) place the general population HIVseropositivity rate for the country at 6070 and for Kampala, the capital city, at 17070. Seropositivity rates exceed 50070 in some young adult groups (7), the age cohort in which most tuberculosis occurs. There are no current estimates of tuberculosis incidence or prevalence in Uganda. Thberculin surveys in 1958 and 1970 found approximately 65070 of adults tuberculin positive with no change between the two surveys (8). In this report, we cite clinical, radiographic, and microbiologic fmdings on 59 patients with pulmonary tuberculosis seen at Old Mulago Hospital, Kampala, Uganda, two-thirds of whom were found to be HIV-seropositive. Fifty-nine adult patients with pulmonary tuberculosis consecutively admitted to Old Mulago Hospital between November 1988 and May 1989 were studied. All patients admitted to the hospital were
SUMMARY The clinical, radiographic, and microbiologic features of 59 patients with pulmonary tuberculosis In Kampala, Uganda were studied and correlated with the serologic reactivity to the human Immunodeficiency virus (HIV) of these patients. 'TWo-thirds of the patients with tuberculosis were HIY-seroposltlve. Histories of fever and weight loss were more prominent In HIY-seroposltlve patients, and perlhllar and basilar Infiltrative diseases were more frequently seen In HIY-seroposltlve patients. Although all patients responded similarly to drug therapy, cutaneous drug reactions were seen In nearly one-third of HIY-seroposltlve patients receiving thiacetazone. AM REV RESPIR DIS 1991; 143:185-187
included in the study if acid-fast bacilli were seen on smear or Mycobacterium tuberculosis was recovered in the culture of at least one sputum sample. One patient with pulmonary disease consistent with active tuberculosis in whom the diagnosis was confirmed by histologic examination of a scrofulous lymph node was also included. Clinical assessment was based upon history and physical examination by one of the investigators (AO). All patients except for one woman in the first trimester of pregnancy had posteroanterior chest radiographs that were reviewed by two of the investigators (PPE and AO). Radiographs were classified as typical if the disease was predominantly an upper lobe disease of a fibronodular, exudative, or cavitary nature and as not typical if the disease was predominantly perihilar, including hilar adenopathy, or basilar and of an infiltrative nature. Expectorated sputum specimens were examined at the National Tuberculosis Reference Laboratory at Wandageya. Sputa were digested with 4010 sodium hydroxide, followed by neutralization with hydrochloric acid. After centrifugation, a sample of the pellet was stained with auramine 0 and examined by fluorescence microscopy. The remainder of the pellet was cultured on a Lowenstein-Jensen
medium. All isolates were tested for niacin production using niacin test strips (Difco Laboratories, Detroit, MI). HIV serologic status was assessed by an ELISA of serum antibody using Wellcozyme anti-HTLV III kits (Wellcome Diagnostics, Dartford, England) or recombigen-HIV EIA kits (Cambridge BioScience, Worcester, MA). All positive HIV ELISA serologies and a 10% random sample of negative serologies were confirmed by western immunoblotting using the novapath immunoblot assay (Biorad Laboratories, Hercules, CAl.
(Received in original form June 8, 1990 and in revised form August 20, 1990) 1 From the Ministry of Health, Thberculosis Control Programme, and Makerere University, Kampala, Uganda and from Case Western Reserve University, Cleveland, Ohio. 2 Supported in part by Grant No. AI -26482 from the National Institute of Allergy and Infectious Diseases, National Institutes of Health. 3 Correspondence and requests for reprints should be addressed to Thomas M. Daniel, M.D., Department of Medicine, University Hospitals, 2074 Abington Road, Cleveland, OH 22106.
