INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Mycobacterium tuberculosis (MT) causes an estimated 3 million deaths annually worldwide, exceeding those caused by any other infectious agent (1). Alleviation of the human and economic costs inflicted by MT requires greater knowledge of the pathogenesis of tuberculosis. The clinical manifestations of tuberculosis depend on the cellular immune responses to the tubercle bacilli, which are characterized by the accumulation of monocytes/macrophages, lymphocytes, and polymorphonuclear leukocytes (PMN) in tuberculous lesions. The immunologic responses are initiated after sensitization of T lymphocytes by the bacterial antigens of MT, with the release of several cytokines that regulate cell functions. The cellular immune response to MT takes the classic form of delayed-type hypersensitivity that is thought to be mediated by a cooperative interaction between T lymphocytes and macrophages, depending upon the interplay of cytokines produced by a variety of mononuclear cells including T cells (2). Recent outbreaks of pulmonary tuberculosis underscore the need for novel approaches to treating and preventing tuberculosis, including the use of immunotherapeutic modalities that enhance normal antimycobacterial defenses. The lung is the only internal organ to which curative or modulating substances can be applied directly through a natural route so that they can exert their influence only in the target organ without affecting the whole organism. Interferon- (IFN-) is a pleiotrophic cytokine with useful infection-control activities; it is produced primarily by mononuclear phagocytes that are stimulated by bacteria and viruses, and which exert a wide range of immunomodulatory activities. The availability of pure IFN- preparations permits administration of this modulator by aerosol. Recent data strongly suggest that IFN- works with interleukin-12 (IL-12) and plays an important role in differentiation toward the T-helper-cell type 1 (Th1) response, inducing IL-2 and interferon- (IFN-), and inhibiting the T-helper-cell type 2 (Th2) compartment (3). It can now be accepted that IFN-, released in large amounts by antigen-presenting macrophages, may favor the production of Th1-like cells by activating the gene for IFN- in CD4⁺ T cells and by antagonizing the effect of IL-4, IL-10, or other immunosuppressive cytokines. Recently, IFN- has been reported to be effective in inhibiting both antigen (Ag)-induced proliferation and cytokine production by Th2 clones, suggesting that IFN- may favor recovery in patients with pulmonary tuberculosis (3, 6). Previous work found that aerosolized IFN- administration to the lung was well tolerated at biologically active doses. 2'-5' Olygoadenilate synthetase (OAS) levels, used as a marker of IFN- activity, increased after local treatment in serum and bronchoalveolar lavage fluid (BALF), suggesting that the inhaled cytokine was active at lung level (9, 10).

The aim of the present study was to evaluate the modifications of biologic (sputum MT concentration, cellular spectrum, and numbers of cells in subpopulations in BALF), clinical (fever and high-resolution computed tomographic [HRCT]), and biochemical (10 cytokines, anti-IFN- antibodies, and IFN- receptors) variables analyzed in subjects with untreated, active pulmonary tuberculosis observed before and after 2 mo of therapy consisting of conventional antimycobacterial treatment with or without aerosolized IFN-.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Population

This was an open, parallel study. The study population consisted of 20 human immunodeficiency virus (HIV)-negative patients with newly diagnosed, active pulmonary tuberculosis who presented with a history and chest radiographic findings compatible with pulmonary tuberculosis and who were enrolled in the Chest Clinic of the Lazzaro Spallanzani Institute in Rome. A sputum smear positive for acid-fast bacilli, fever, and loss of weight before admission to the hospital were criteria for inclusion. In addition, all of the patients had positive tuberculin skin tests and no history or signs of any other medical condition requiring current treatment. Bactec (Bactec 460 TB; Becton Dickinson, Milan, Italy) analysis and sputum culture for MT confirmed its presence. Sensitivity tests were adopted to exclude multiresistant strains of MT. Following the results of the susceptibility testing, two patients were excluded from the study because they had drug-resistant strains of MT. All patients were mantained as inpatients, and therapy was administered only by the professional staff.

Patients were randomly subdivided into two groups. Group A (n = 10, comprising six females and four males; three smokers and seven nonsmokers; mean age = 31 ± 7 yr) was given antituberculous chemotherapy consisting of isoniazid (INH; 5 mg/kg daily), rifampicin (RAMP; 10 mg/kg daily), ethambutol (ETB; 15 to 25 mg/kg daily), and pyrazinamide (PZ; 15 to 30 mg/kg daily). Group B (n = 10, comprising five females and five males; four smokers and six nonsmokers; mean age = 34 ± 13 yr) received a combination of the same antituberculous drugs plus aerosolized human lymphoblastoid IFN- (HuIFN-Ly; 3 MU/dose three times weekly). The dosing schedule was continued for 2 mo. Fever was assessed every day until resolution, and weight was checked at 1-wk intervals.

Fiberoptic bronchoscopy with bronchoalveolar lavage (BAL) was performed 24 h after baseline spirometry. The BALF was subjected to cytologic, analysis and assay for anti-IFN- antibodies and concentrations of several cytokines. From 2 to 4 h after the last dose (total study dose = 72 MU/2 mo) of HuIFN-Ly, all of the baseline measurements mentioned previously were repeated. Blood samples were taken from the patients before and after treatment. Chest radiography and HRCT were done after 2 mo. None of the patients was given corticosteroids or other immunosuppressive agents. The study was approved by the hospital ethics committee, and all subjects gave informed consent.

Safety of Treatment

Patients in Groups A and B were monitored for local and systemic signs and symptoms, and were subjected to routine hematologic blood sampling and blood chemistry and coagulation analyses as well as urinalysis and pulmonary function testing (MasterLab Transfer Spirometer; Jaeger, Würzburg, Germany).

Sputum Induction Protocol and Conversion for MT

Sputum was collected from all subjects. Direct microscopy of smears prepared from sputa after sodium hypochlorite treatment and centrifugation was done every week for 2 mo to evaluate the concentrations of tubercle bacilli. To produce a sputum specimen of sufficient quality and quantity (l to 2 ml), patients inhaled a mixture of nebulized 0.9% hypertonic saline delivered by an Ultraneb 99 Ultrasonic nebulizer (DeVilbiss, Langen, Germany) for 10 min in a total volume of 15 ml of hypertonic saline. Sputum was transferred to a 10 ml screw-cup tube. An equal volume of undiluted sodium hypochlorite (5%) was added and the mixture was incubated at room temperature for 10 min. The tube was shaken regularly. Distilled water was added to fill the tube, and the sample was centrifuged at 3,000 g for 15 min. The sediment was used to prepare smears stained with the Ziehl-Neelsen method. The number of MT was expressed as the mean number MT in 20 fields of view by microscopy (magnification ×40) (11).

HRCT Evaluation

HRCT studies were done with an A-TOM XR 6000 (Ansaldo Elettronica Biomedicale, Genoa, Italy). Sections 1.5-mm thick at 10-mm intervals, with a 512 × 512-pixel reconstruction matrix, instrument settings of 130 kV and 175 mA, and a high-spatial-frequency algorithm were used. All images were obtained at the suspended end- inspiratory volume with a 1.9-s scanning time. All patients underwent scanning in the supine position. All images were obtained at window settings appropriate for lung parenchyma (level = 400 to 800 HU; width = 1,800 to 1,600 HU). HRCT scans were evaluated for the presence, distribution, and extent of the following signs of pulmonary tuberculosis: (1) miliary nodules (1- to 2-mm nodules involving both the intralobular interstitium and interlobular septa); (2) nodules ( 10-mm nodules related to the terminal or respiratory bronchioles and separated from the pleural surface or interlobular septa by more than 2 mm); (3) consolidation (panlobular and polilobular consolidation); (4) a ground-glass appearance; (5) cavitation; (6) bronchial lesions (bronchial wall thickening, bronchiectasis), and (7) fibrotic bands. The extent of involvement was assessed for each of the three zones of the lung. On the HRCT scans, the lungs were divided into six zones (upper, middle, and lower zones of the right and left lungs). The upper zones were defined as areas of the lung above the level of the carina; the middle zones as areas between the level of the carina and the level of the inferior pulmonary veins; and the lower zones as areas below the level of the inferior pulmonary veins. The HRCT score in upper, middle, and lower lung zones was determined by visually estimating the extent of disease in each zone (12). Retrospectively, the HRCT scans were blindly reanalyzed by two independent chest radiologists, and final conclusions on the findings were reached by means of consensus. The HRCT score was based on the percentage of lung parenchyma that showed evidence of each recorded abnormality. A score of 1 reflected involvement of less than 25% of the image; a score of 2, involvement of 25 to 50%; a score of 3, involvement of 50 to 75%; and a score of 4, involvement of more than 75%. The scores for each zone were then added to obtain a global extent score, ranging from 0 to 24, referred to as the HRCT extent score of each HRCT abnormality. A total score of lung involvement was obtained by summing the global extent scores of all HRCT abnormalities, ranging from 0 to 192, as an expression of disease severity.

Processing of BALF

Fiberoptic bronchoscopy was performed before and after antituberculous chemotherapy in all cases. The baseline bronchoscopy was done within 24 h before the initiation of IFN- treatment. The final bronchoscopy was performed after the last dose of aerosolized IFN-. Bronchoscopy with BALF collection was done according to a previously described method (13). Prebronchoscopy medications for each patient included intramuscolar injections of 0.5 mg of atropine; 4% xylocaine nebulized into the nasopharynx; a total of 4 ml of 1% xylocaine applied directly to the laryngeal area via the bronchoscope, and 2 ml of 1% xylocaine applied directly to the trachea. The bronchoscope was inserted near a segment of the affected lung with chest radiographic guidance. A standard lavage protocol was followed by infusing five aliquots of 20 ml each (a total of 100 ml) of sterile 0.9% saline, at body temperature, through the aspiration port of the bronchoscope. Lavage fluid was collected via the same port after infusion of each aliquot by gentle suction into a plastic trap. The fluid recovered was filtered through sterile gauze and the volume was measured. BALF analysis was performed immediately after bronchoscopy, and the results were independently verified by a single technician blinded to the patient's condition. The first aliquot was used for counting total cells with a Burker hemocytometer, and the differential cell analysis was done by counting 400 cells in smears stained with May-Grunwald-Giemsa. The lymphocyte subpopulation analysis was done flow cytometrically, using a second aliquot of BALF (with kits purchased from Becton-Dickinson, Mountain View, CA). The remaining aliquot was centrifuged at 1,200 × g for 10 min. The cells were washed and resuspended in RPMI (2.5 × 10⁶ cells/ml), and supernatants were frozen at 80° C until used.

Measurement of Cytokines in BALF

The epithelial lining fluid (ELF) dilution was determined by the urea method (14) on BALF samples. To determine the urea content of lavage fluid samples, a commercially available kit (Sigma 65 UV; Milan, Italy) was purchased from Sigma. Cytokine concentrations were measured in supernatants of BALF after 20-fold concentration by means of Amicon concentrators (Amicon, Beverly, MA). Human IL-1, IL-2, IL-4, IL-6, IL-8, IL-10, IFN-, and TNF- were measured with an enzyme-linked immunosorbent assay (ELISA; R&D Systems, Minneapolis, MN). Levels of IFN- were measured with a radio immunoassay (RIA; Nuclear Laser Medicine, Milan, Italy). IL-12 was measured with an IL-12 + p40 enzyme amplified sensitivity immunoassay (EASIA) kit (Medgenix Diagnostics, Fleurus, Belgium). Kit sensitivities were as follows: IL-1 = 0.3 pg/ml; IL-2 = 2.5 pg/ml; IL-4 = 4.1 pg/ml; IL-6 = 0.7 pg/ml; IL-8 = 18 pg/ml; IL-10 = 2 pg/ml; IL-12 = 1.5 pg/ml; IFN- = 0.3 pg/ml; IFN- = 3 pg/ml; and TNF- = 4.4 pg/ml.

Expression of IFN- Receptors

An aliquot of BALF cell preparation was incubated at 4° C. The cells (2.5 to 5 × 10⁵) were added to a final volume of 50 µl/well and washed by adding 150 µl of phosphate-buffered saline (PBS)/NaN₃ to the wells and centrifuging the cells at 1,000 × g for 3 min. The wash buffer was subsequently aspirated. Cell permeability was obtained by adding and evaporating methanol (10° C) for 5 min. Cells were again washed and incubated with 50 µl of anti-IFN- receptor (IFN-R [R-100] antibody; Santa Cruz Biotechnology, Santa Cruz, CA) for 20 min. The optimal antibody concentration was chosen according to the manufacturer's guidelines. Subsequently, the cells were incubated with 50 µl of fluoresceinated goat F(ab)₂ antirabbit total Ig (Santa Cruz Biotechnology) for 20 min at 4° C. The cells were washed and fixed by adding 50 µl of a 2% solution of paraformaldehyde (pH 7.4) and were incubated at 4° C for 5 min. Following this, the cells were washed to remove residual paraformaldehyde and unbound fluorescent antibodies. The cells were then analyzed flow cytometrically with a FACScan cell sorter (Becton Dickinson). The positive fluorescence intensities were referred to the negative control. For both macrophages and lymphocytes, the results were expressed as relative IFN-R log fluorescence intensity and percentage of IFN-R-positive cells.

Serum Determination of anti-IFN- Antibodies

Anti-IFN- antibodies were measured with an ELISA (Nuclear Laser Medicine).

Aerosolized Administration of IFN-

HuIFN-Ly, kindly furnished by Glaxo-Wellcome (Verona, Italy), was given by an IS-2 jet nebulizer operating on compressed air (Pari, Starnberg, Germany), generating a total output of 0.32 g/min at an inspiratory flow of 20 L/min, with a delivery time of 4 min and delivery volume of 1 ml. Particle size analysis was done with a high-resolution analyzer (Aerosizer Mach2; API, Milan, Italy). Approximately 90% of the aerosol droplets produced ranged between 0.40 and 3.4 µm mass median aerodynamic diameter (median = 1.374 µm; standard geometrical deviation = 1.340 µm [respirable range]). The International Committee on Radiation Protection has fixed a value of 30% deposition in the lower respiratory tract for droplets this size. Therefore the amounts of IFN- administered were not comparable to those received. Throughout this report, the administered dose is defined as equivalent to the aerosolized dose.

Statistical Analysis

Owing to the limited number of patients examined, to the lack of knowledge of the data distributions, and to the inevaluability of a number of samples for some of the variables considered, we used a nonparametric statistical analysis. Therefore, the results were expressed as medians and ranges, and comparisons were made by using Wilcoxon's paired rank-sum test for paired measurements or the Mann-Whitney U test for unpaired measurements. To analyze the differences in the data at baseline (T0) and 2 mo (T1) (see Table 1) in both Groups A and B, the T0-T1 differences in the variables were compared, using the Mann-Whitney U test.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Biological Variables

BALF cellularity. The general characteristics of BALFs are presented in Table 1. The total volume of fluid retrieved, the total number of cells, and the absolute numbers of macrophages and neutrophils per milliliter observed before treatment were not significantly different from those found after therapy in both study groups. By contrast, the lymphocyte counts in Group B were significantly increased after treatment (p = 0.02). No differences were observed as far as the number of lymphocyte alveolitis cases is concerned in the two groups.

The CD4⁺ and CD8⁺ T-cell counts were not significantly modified after as compared with before treatment in either study group. Group B, however, showed a significant difference (p = 0.02) in HLA-DR⁺ T cells after as compared with before treatment. An analysis of the differences (pre- minus posttreatment values) was done for all variables considered. A significant difference was noted only for HLA-DR⁺ T cells (p = 0.006).

Sputum conversion for MT. The mean numbers of MT bacilli counted in sputum smears (Figure 1a) showed that IFN--treated patients (Group B) had significantly reduced counts of bacilli at the end of the first week of treatment, whereas the counts in the group treated only with antituberculous chemotherapy (Group A) changed significantly only at the end of the second week of therapy. The analysis of the changes for each patient (pre- minus posttreatment values) showed a significant decrease in numbers of bacilli at the end of the first week (p = 0.04).

View larger version (23K): [in this window] [in a new window]   Figure 1.   (a) Rates of sputum microscopic conversion to Mycobacterium tuberculosis in two groups of patients. Closed squares: patients treated for 2 mo with RAMP + INH + ETB + PZ; open squares: patients treated for 2 mo with RAMP + INH + ETB + PZ + IFN-. (b) Median levels of fever measured in Groups A and B of tuberculosis patients. Solid lines: Patients treated for 2 mo with RAMP + INH + ETB + PZ; broken line: patients treated for 2 mo with RAMP + INH + ETB + PZ + IFN-.

Clinical Findings

Vital signs. Neither Group A nor Group B showed significant post- versus pretreatment changes in vital signs (blood pressure, heart rate, respiratory rate, and weight; data not shown).

Adverse effects. All patients tolerated the therapy well. No adverse effects were registered in the group of IFN--treated patients.

Fever. Figure 1b shows that fever in the two groups behaved differently during the first 6 d of treatment. In fact, in evaluating the changes in temperature for each patient (pre- minus posttreatment values), a significant difference was observed in the two groups at the third (p = 0.049) and fourth (p = 0.02) days after the beginning of therapy.

Evolution of HRCT findings. The reversibility of pulmonary damage as a result of 2 mo of treatment was evaluated. In Group A, the overall disease scores on initial HRCT scans ranged from 10 to 85 (median = 31), whereas the overall disease scores on follow-up scans ranged from 2 to 72 (median = 24; p = 0.025). In Group B, the overall disease scores on initial HRCT scans ranged from 16 to 94 (median = 29), whereas on follow-up scans the overall scores ranged from 7 to 75 (median = 19; p = 0.009). Data for the frequency and extent of HRCT abnormalities at the first and second evaluations are given in Table 2 for each HRCT finding. In Group A, a decrease in the HRCT scores for the frequency and extent of specific abnormalities was found after 2 mo of therapy, but the difference was not statistically significant. In Group B, in which the patients underwent adjunctive treatment with aerosolized HuIFN-Ly, HRCT abnormalities at the second evaluation showed statistically significant differences from those at the first evaluation for nodules (p = 0.028) and consolidation (p = 0.009) scores.

Biochemical Variables

Cytokines in BALF supernatants. The release of cytokines into BALF supernatants from the involved lung segments of tuberculous patients, as measured with ELISA, is shown in Table 3. After 2 mo of treatment, none of the cytokines assigned reached levels statistically significantly different from the levels observed before treatment, with the exception of TNF- and IL-6 in both Groups A and B; IL-1 was significantly decreased only in Group B. However, it is important to note that the differences were more significant in Group B (Table 3). When the T0-T1 changes were compared in Groups A versus B, TNF- and IL-6 were still significantly different. New findings were significant differences for IL-2 and IFN-. Their behavior was consistent with that of the CD4/CD8 T-cell ratio shown in Table 1.

IFN- receptor levels. Figure 2 shows IFN-R expression as evaluated flow cytometrically in BALF cells of Groups A and B before and after treatment. In Group A, lymphocytes evaluated after therapy had a median log fluorescence intensity not different from that before therapy (with 85% of the cells showing specific IFN-R positivity before and 77% after therapy). Before therapy in Group B, the lymphocytes had a median log fluorescence intensity significantly lower than that observed after therapy (p = 0.01) (the number of cells showing specific IFN-R positivity rose from 73% to 75%).

View larger version (18K): [in this window] [in a new window]   Figure 2.   Changes in BALF IFN-R levels following 2 mo of antituberculous chemotherapy in Groups A and B of patients. T0: pre-therapy; T1: posttherapy; Group A: patients given antitubercular therapy (RAMP + INH + ETB + PZ); Group B: patients given same antitubercular therapy plus aerosolized IFN-. For each patient, the IFN-R log fluorescence (Fluor) intensity obtained through flow cytometry is reported both for lymphocytes (open circles) and for macrophages (closed circles). Horizontal bars indicate median values; p represents statistical significance.

In Group A, macrophages showed a median IFN-R log fluorescence intensity that did not change statistically from before to after treatment. Before treatment, 94% of the cells showed specific IFN-R positivity, which increased to 97% after treatment. The same results were observed for macrophages in Group B (with 93% of the cells showing specific IFN-R positivity before therapy and 91% after therapy).

Serum determinations of anti-IFN- antibodies. The assay for anti-IFN- antibodies was negative in all patients, before and after 2 mo of treatment (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pulmonary tuberculosis is an infectious disease caused by MT, a very resistant microorganism with intracellular localization. At present, antimycobacterial therapy includes a panel of combined antibiotics intended to provide maximum efficacy and rapidity of disease resolution and limitation of side effects. Despite this massive therapeutic approach, treatment must be continued for a long time (at least 6 mo) to avoid frequent relapses caused by failure to eradicate the microorganism in the short term (15). Any therapeutic approach that would enhance natural host defenses would clearly be important once its efficacy was established. Different immunomodulatory substances have been reported as capable of controlling the infection with MT. Generally, these substances belong to the group of cytokines, especially IL-2, IL-12, IFN-, and TNF-. These molecules exert their effects by activating different types of cells involved in microorganism clearance (16). Another cytokine molecule, IFN-, has recently been reported to be capable of controlling MT (20). However, this cytokine produces toxicity, and we therefore decided to instead explore the effects of IFN-, which is less expensive and has been more extensively studied in humans (21, 22).

Originally considered a simple antiviral substance, IFN- has since been shown to exhibit a variety of biologic effects. Early studies reported multiple effects of IFN- on the immune response in vitro as well as in vivo, including effects on antibody production, natural killer (NK) cell activity, antigen presentation, and phagocytosis (23). More recently, IFN- has been shown to regulate T-cell function and to particularly stimulate the Th1 subset. In contrast, it inhibited both antigen-induced proliferation and cytokine production by Th2 clones (6). Consequently, IFN- can increase the IFN--producing CD4⁺ population of T-cells and antagonize the suppressing effects of IL-4 (3). In mice, cytokine production (such as IL-2 and IFN-) by Th1 cells is associated with protection against tuberculosis, whereas Th2-produced cytokines are associated with increased susceptibility to the disease (24, 25). Furthermore, animal and in vivo studies also show that progressive tuberculosis is associated with a Th2 response, whereas regression of the infection is characterized by enhancement of the Th1 compartment and release of Th1-produced cytokines (26). Therefore, one of the aims of the present study was to evaluate the effect of IFN- administration in combination with conventional chemotherapy in patients with pulmonary tuberculosis.

To avoid the side effects already described for IFN-, we followed the direction provided by a previous study done by our group (10) and decided to treat patients with IFN- given by aerosol. As previously shown, this cytokine can be safely delivered to the epithelial surface of the lower respiratory tract without systemic side effects (10). To establish possible mechanisms involved in infection control by IFN-, we evaluated several biologic, clinical, and biochemical variables. Among these variables we found that the number of MT in sputum and the pattern of cells in BALF were significantly modified when pre- and posttreatment determinations were compared (Figure 1 and Table 1). In fact, sputum microscopic conversion was clearly anticipated by a week after the beginning of treatment in patients receiving the combination therapy including IFN-. Since different cell types (such as monocytes/macrophages, T lymphocytes, and neutrophils) are considered important mediators of bacterial clearance, we analyzed BALF cells both qualitatively and quantitatively by means of flow cytometric techniques. Pretreatment findings in the patients on whom this was done were consistent with those reported in the literature, with increased numbers of lymphocytes and granulocytes, as well as decreased numbers of monocytes (27). When we made a comparison of flow cytometric results before and after 2 mo of therapy in our study Groups A and B, only Group B, receiving IFN-, showed significant increases in total lymphocytes and HLA DR⁺ cells. The total lymphocyte increase agrees with the known protective effect of CD4⁺ T cells against Mycobacterium avium in mice (28). Besides evaluating biologic variables, we evaluated two clinical aspects of tuberculosis: fever and radiologic features. Corresponding to MT reduction in sputa, fever also decreased more rapidly in patients receiving IFN-. This is a second important indication that combination therapy including IFN- may have a rational role in treating tuberculosis.

The third crucial indicator used to follow the evolution of pulmonary tuberculosis was HRCT. More sensitive than chest radiography for the detection and characterization of subtle parenchymal disease, HRCT may also be of value in detecting and following diffuse lung involvement (when corresponding chest radiographs are normal), or in showing minimal or limited disease (29). In our study, HRCT permitted the evaluation of some of the specific variables (listed in Table 2) that can describe the evolution of tuberculosis (30). These include two abnormalities: nodules and areas of consolidation. HRCT scores represent both the most important variables of activity for pulmonary tuberculosis and those variables showing the most rapid variations over time. Therefore, the significant reduction in scores for these variables, observed only in IFN--treated patients, could corroborate the utility of IFN- treatment. As far as biochemical variables are concerned, several determinations were done. As is known, cytokine receptors are generally induced by the same molecule they bind (31), and their increase therefore represents a marker of cytokine activity. Since the treatment in one arm of our study was based on combination therapy including IFN-, we analyzed IFN- receptors both on lymphocytes and on monocytes/macrophages from BALF. The data showed that IFN- receptors expressed on BALF lymphocytes of patients receiving IFN- were increased concomitantly with disease improvement. In the same period, measurements of three cytokines showed significant decreases from before to after 2 mo of therapy. In particular, levels of IL-1, IL-6, and TNF- dropped in BALF after treatment.

It was also noted that in patients not given IFN-, the same cytokines also fell to lower levels than before treatment, although significant differences were observed only for IL-6 and TNF-. Furthermore, the significance of the cytokine variations was greater in IFN--treated than in non-IFN--treated patients. This suggests that changes in proinflammatory cytokines, possibly linked to the reduction of inflammatory phenomena, are associated with disease improvement in tuberculosis. It is not surprising that the three cytokines named here, IL-1, IL-6, and TNF-, exhibited behavior according with this finding. These three cytokines act synergistically, and their amounts often correlate in different diseases. A more sensitive comparison of the T0-T1 (see Table 3) changes observed for the different cytokines in Groups A and B revealed possibly important differences in IL-2 and IFN-. Literature data show that TNF-, IFN-, and IL-2 can exert antimycobacterial activity (18, 32). The approach used in our study did not permit a specific analysis of which kind of cells produced these mediators, although it is reported in the literature that a wide spectrum of cells can release them, especially activated monocytes/macrophages and T lymphocytes. Our findings did not reflect increases of IFN- in IFN--treated patients, possibily because of the short, 2-mo posttreatment observation time, during which both the effect of IFN- and disease recovery were effective. A particular point of note is the methodology we used for determining and expressing cytokine levels. In our study, cytokine concentrations were normalized by the dilution factor determined by the volume of the saline injection used to obtain BALF (14). The few data reported in the literature on this topic generally show BALF cell cytokine release, making comparisons with our results impossible (33). In addition, no data are available on comparisons of cytokine concentrations analyzed before and after therapy.

In conclusion, even though the data discussed do not clarify the mechanisms involved, the patients receiving conventional chemotherapy plus IFN- in our study showed more favorable effects than those given chemotherapy alone. The data presented here, if confirmed with a greater number of subjects, open a wide spectrum of possibilities, including that of using aerosolized IFN- in both HIV-negative and HIV-positive tuberculosis patients to enhance their natural response against MT. Specific trials must be planned to explore the best schedule and doses for such treatment, and the possibility that IFN- favors the clinical outcome in patients infected with multiresistant strains of MT also needs to be evaluated, as does the possibility that shorter periods of conventional therapy may be established to limit side effects. In addition, experiments in vitro and in animal models are needed to analyze the possible mechanisms involved in the effects of treating tuberculosis with IFN-.