The prevalence of isolated atypical Mycobacteria (also called non-tuberculous Mycobacteria) infection is growing and is suspected to surpass Mycobacterium tuberculosis in some areas. Mycobacterium avium complex (MAC) species are the most frequent. Immunocompetent patients develop lung disease as a progressive chronic pneumonia. In the United States and in Japan there are approximately 1.3 cases per 100 000 persons, whereas in Switzerland, there are 0.9 cases per 100 000 persons with pulmonary MAC. It is postulated that up to 70% of patients with pulmonary MAC infection had a previous bronchopulmonary disorder 1. However, cases without a predisposed previous condition or identifiable immunodeficiency also exist 2.

Atypical Mycobacteria can survive and proliferate within the hostile environment of host macrophages 34. The treatment of these pulmonary infections is more difficult than Mycobacterium tuberculosis due to a higher drug resistance. MAC and Mycobacterium fortuitum-cheloneae are the two most resistant. The problem to choose an adequate antibiotic combination derives from the lack of controlled, prospective studies. The different populations, bacteriological methods and cure approaches applied, as well as lack of correlation between clinical and laboratory results add higher complexity. Conventional chemotherapy is prolonged and associated to frequent toxicities 567. Mortality attributed directly to infection occurs, but its rate is ignored.

Therefore, an alternative therapeutic target can be directed to the manipulation of the host's defenses. As most of the intracellular infections, the immune response to control infection depends primarily on a Th1 lymphocytes outbreak 8. Interferon (IFN) gamma, a glycoprotein produced primary by CD4+ cells, plays a main role in this process. Enough evidences exist related to the action of IFN gamma on the immunoregulatory activity of macrophages 9. Lack of production of this cytokine 10 or expression of its receptor 111213 is associated to the infection's most lethal forms. IFN gamma has also a potent antifibrotic effect 1415.

Previous clinical studies have demonstrated satisfactory results with the use of recombinant IFN gamma in patients infected with atypical Mycobacteria 1617181920. However, none of these reports was a controlled trial and they included only few patients. The present placebo-controlled study was done with the objective to assess the immunoadjuvant IFN gamma effect in patients with pulmonary atypical Mycobacteriosis regarding their clinical, bacteriological and radiological evolutions. Additionally, several immune response and oxidative stress markers were measured. The safety of the procedure was also evaluated.

A randomized, double-blind, placebo-controlled trial was carried out at the "Benéfico Jurídico" Hospital, Havana, which is the national reference unit for tuberculosis and other respiratory diseases and where most of the patients with unfavorable evolution are remitted. Another participant institution was the "Amalia Simoni" Hospital, Camagüey. The clinical protocol was approved by the Ethics Committees of the corresponding sites and by the Cuban Regulatory Authority. The study complied with the Declaration of Helsinki.

Thirty-two patients were enrolled from October 2002 to March 2005; then follow-up continued up to September 2005. Eighteen patients received IFN gamma and 14 placebo. Most of the patients (84.4%) corresponded to "Benéfico Jurídico" Hospital. Table 1 summarizes the course of the trial. Sixteen patients (88.9%) completed the 6-months treatment period in the IFN group (two abandoned after the first month), while 10 (71.4%) completed it in the placebo group. Withdrawals in this latter group were 2 deaths after the first month (MAC infection and pulmonary thromboembolism) and 2 voluntary abandons after months 2 and 5. The patient that abandoned after the second month of treatment died 2 months later due to a lung cancer, not present at entry or at withdrawal. During follow-up 2 patients died in the placebo group (myocardial infarction and pulmonary thromboembolism) and other 2 in the IFN group (both myocardial infarctions). In addition, one IFN patient did not attend the last evaluation. Finally, 13 patients (72.2%) completed the entire study in the IFN group, and 8 (57.1%) in the placebo group.

Baseline characteristics of the included patients are shown in Tables 2 and 3. Groups were homogeneous. Even if there were noteworthy differences concerning smoking and alcohol drinking habits, the sample size does not provide enough power to assess their relation to response. Mean age was around 60 years for both groups and white men prevailed (> 70%). M. avium-intracellulare complex was 94% of the isolated species. None of the respiratory antecedents stood out particularly. All the patients presented breathing disorders (movements, crepitations) at physical examination, but none had any extra-pulmonary manifestation of the disease.

The results of the evaluations are shown in Table 3. The overall response, which was the main study outcome, was significantly better in those patients treated with the combination (72% vs. 36%). All responses were complete, except for one patient in the IFN group who responded partially after treatment but reached complete response at the end of follow-up. The difference persisted during the follow-up (67% vs.29%). These analyses were done under the intention-to-treat principle, where all withdrawals are considered as failures. If the analysis is made with each patient's last evaluation differences are larger and more significant.

The clinical, radiological, and bacteriological assessments taken independently were also better for the IFN gamma group. Improvement was obtained for the presence of respiratory symptoms such as dyspnea only in the IFN group. Additionally, withdrawals in this group were also asymptomatic, which only happened for one of the placebos. BMI increments were also larger among IFN treated patients. Regarding clinical general status and its improvement, statically significant differences were achieved. Importantly, 7/14 patients from the IFN group but none among the controls improved their general status (passed from "bad" to "moderate" or from "moderate" to "good") just one month after the therapy onset. Among survivors, there were two patients in the placebo group who worsened their general clinical condition. The only patient where this happened in the IFN group was infected with M. fortuitum-cheloneae.

All patients, except one in the IFN group, presented fibrotic lesions. Twelve individuals in each group had cavitations. Pleural thickness and interstitial and/or alveolar infiltrate were seen in 15 and 13 subjects, respectively. Apical and parahilar lesion localization were most common. Advanced lesions predominated at entry in both groups. At the end of treatment more patients treated with IFN gamma reduced lesion extension: advanced lesions persisted in only 3 patients from this group vs. more than half in the placebo group. A 61 years-old male, 2 years with the disease, which initially had advanced lesions evidenced a total disappearance of lesions after IFN treatment and follow-up (Figure 1). The other 12 patients in the IFN group who where evaluated at month 18 improved their radiological condition, 5 of them reached the minimum lesion extension and 4 up to moderate. On the contrary, among the controls, 6 improved, but only one reached minimum extension, and it was a subject bearing M. kansasii infection. Additionally, one patient in this group worsened his status at 18 months follow-up. Cavitary lesions disappeared in 5 patients in the IFN group (42%), at an average of 6 months (range: 1 – 12 months), but only in the M. kansassi bearing patient in the placebo group (8%), at 5 months.

Bacteriological response by direct examination and culture occurred in both groups at the end of treatment. However, during follow-up there were more relapses in the placebo group (3 vs. 1 in the experimental group) and besides, there was one late positive sputum conversion to negative among the IFN receiving patients. Statistically significant differences were attained regarding sputum culture codification at the end of follow-up.

Overall complete response was obtained earlier in patients treated with the combination (Figure 2; log rank test, p = 0.024). Interestingly, the only complete responder at 1st month of treatment corresponded to the IFN group. All symptoms disappearance, or particularly dyspnea, and radiological improvement behaved the same (graphs not shown). The most important differences in improvement rate concerned the radiological evaluation. After one month of treatment, 4 patients in the IFN group but none among placebos showed pulmonary lesions improvement. At this point, 60% of patients with initial dyspnea in the IFN group, no longer presented this symptom, in contrast to16.7% in the other group.

Survival profiles are shown in Figure 3. During the entire study, 5 patients (35.7%) deceased in the control group (median survival: 4.5 months) and only 2 (11.1%) in the IFN group. Both deaths in the IFN group occurred at 16 months, as complete responders. Among the control patients, 4 died as non-responders and the other one as partial responder.

Oxidative stress parameters basal mean values were importantly increased in the patients as compared to a group of 60 Cuban healthy individuals 26 (Table 4). Significant decrements in SOD and AOPP values were detected in the IFN group (Table 5). The difference was more pronounced for AOPP, since it decreased in 9/10 IFN treated patients whereas increased in 5/7 from the placebo group. For the other parameters important changes were not perceived.

Regarding gene modulation, the most important variations were detected within the groups after treatment (Table 5). TGF beta increased significantly in the placebo group, whereas MCP-1 decreased in the IFN gamma group. TNF alpha values, although slightly different before treatment, increased significantly in the placebo group, while did not change with IFN therapy. No changes were detected for the other genes evaluated (results not shown).

Seventeen adverse events were presented during the treatment (Table 6). At least one event occurred in 66.7% of the patients in the IFN group and 42.9% in the placebo group. Arthralgias (50.0%), fever (44.4%), chills (33.3%), anorexia (27.8%) and asthenia (22.2%) prevailed in the IFN group, where a higher frequency was detected for fever and chills, which were not present in the placebo group. Most of the events in the IFN group were classified as moderate (60%), none of them severe. All of them were well controlled with or without treatment. Respiratory or cardiorespiratory manifestations, closely related to the disease, were not accounted as adverse reactions.

Significant differences or variations in vital signs and clinical laboratory tests were not detected. Only a moderate creatinine increase and a mild anemia in a single patient of the IFN group were considered as adverse reactions. On the contrary, in this group, 3 patients with complete response normalized their GSR values, 2 of them with initial values > 100 mm/h. A mean reduction of 22 mm/h (71 to 49 mm/h) took place in the IFN group, while in the placebo group GSR values remained stable.

This report constitutes the first and largest randomized, controlled clinical study, using an immunomodulating agent systemically in atypical Mycobacteria infection. The group of patients treated with IFN gamma and antibiotics had a better response from the clinical, radiological and bacteriological points of view. Response rate in the IFN group doubled that from the controls after 6 months of treatment and the difference persisted after one year follow-up. Patients that received IFN gamma responded 5 to 6 months before those with placebo, with respect to all evaluation criteria. The earlier disappearance of dyspnea in the IFN group was remarkable since this symptom affects quality of life more than others. This was not only statistically but also clinically significant. There was also an advantage in overall survival. In fact, except for one case, all patients treated with IFN gamma received benefit from treatment. The 66.7% response rate was calculated under intention-to-treat basis, considering the abandoners and deceased as failures. But in this group one of the abandoners had complete response at withdrawal after the first month of treatment and the two deaths were due to myocardial infarcts. Both deceased cases had had complete response at the end of treatment. In the control group one of the abandoners had responded but four of the five deceased were non responders.

Advanced age and male gender predominated in this study, comparable to reported 1. However, antecedents of respiratory tract affectations were scarce. As expected, MAC species predominated. M. fortuitum-cheloneae was present in one patient of the IFN group, which was the only one that did not receive benefits from treatment. Coincidently, infection by this species has been reported in patients with anti-IFN gamma autoantibodies 27. Nevertheless, infection with this specie is less common. The other species, M. kansassi, presented in one patient with placebo, is more sensitive to first line therapy. This patient was the best responder in this group; the only one where cavitary lesions disappeared. Obviously, differences between groups could be more significant if no-MAC species are excluded from analysis, but this was not previewed in the protocol.

The antibiotic scheme used in this trial was in correspondence with the best update recommended. Based on in vivo and in vitro studies treatment should include: (i) a macrolide (clarithromycin or azithromycin), (ii) ethambutol, (iii) rifampin or rifabutin, and (iv) ciprofloxacin, streptomycin or clofazimine. However, the response level obtained in the IFN group is rarely achieved with antibiotics only. Besides, clinical studies are not controlled, often use more flexible response criteria and longer periods of treatment 28.

Radiological response during treatment was the most significant benefit added by IFN gamma. Improvement was evident since the first month and there was even complete lesion disappearance in one patient, which has not been reported after such a relatively short follow-up time. This lesion extension reduction had consequences on the patients' clinical performance with dyspnea and other symptoms reduction or disappearance. In this case, radiology had a higher clinical impact than bacteriology. Several patients in the placebo group even with sputum conversion, never experienced radiological improvement, most of them worsened lesions and had bacteriological relapse. Contrarily, in the IFN group, some patients with sputum-culture positive had an early lesions reduction, with symptoms disappearance and had later complete response. Improvement difference included not only cavitary lesions reduction (42% vs. 8%) but also fibrosis and infiltrate extension. This suggests a different mechanism for the additional radiological improvement, probably related to the potent antifibrotic effect of IFN gamma, whether the Mycobacterium is present or not.

Treatment was well tolerated. Flu-like symptoms such as fever, chills and arthralgias are among those expected for interferons since their first clinical applications 29. Reduction in blood cell counts and moderate creatinine increase, which appeared in one case each have been also reported. Gastrointestinal events (vomiting, diarrheas) are probably more related to the use of the antibiotics 157. All deaths in the IFN group were during the follow-up period so they are unlikely related to treatment. However deaths can be related direct or indirectly to the disease since it is a predisposing factor for lung cancer, pulmonary thromboembolism, and myocardial infarction, which are all preceded by cardiorespiratory events.

The role of IFN gamma as the main macrophage – activator Th1 cytokine has been clearly established in animal models infected with M. avium 3031. IFN gamma action on the macrophages leads to intracellular Mycobacteria killing. It stimulates macrophages to produce TNF alpha, oxygen free radicals and nitric oxide, increases MHC surface antigens and Fc receptors display, decreases lysosomal pH, and increases the intracellular concentration of some antibiotics (e.g. macrolides) 930. Therefore, its use as adjuvant is justified since existent multidrug therapy, despite its limited efficacy, must be offered to the patients.

Regarding its antifibrotic effect, IFN gamma inhibits lung fibroblast proliferation and chemotaxis in a dose dependent manner and reduces collagen synthesis 14. Furthermore, this protein is a potent antagonist of TGF beta 15, involved directly in severe lung fibrosis progression 32. IFN gamma contributes to the tissue repair and its remodeling 33. These actions were also evident in the results reported with IFN gamma in patients with idiopathic pulmonary fibrosis 3435 and suggests that this cytokine could have future indications in other pulmonary diseases where fibrosis is present, although a definite effect on survival is still to be demonstrated.

Systemic or aerosolized IFN gamma have been reported as satisfactory in other similar intracellular infections 363738394041, but none of these studies was controlled. The concept that IFN gamma can be useful in these atypical Mycobacterial infections is supported by several reports on individuals where the IFN gamma action is impaired. Mutations in the IFN gamma receptor genes are associated with frequent recurrences or the most serious forms of the disease 111213. Patients with defects in the production of IFN gamma or other cytokines or partial deficiencies of IFN gamma receptor can obtain benefits with IFN gamma treatment 101342. Additionally, patients without genetic disorders but with serum anti-IFN gamma autoantibodies have a higher susceptibility to develop Mycobacteriosis 43.

IFN gamma has been effective as adjuvant in AIDS patients co-infected with MAC, where a clear decrement in the bacteremia was verified. These results were obtained in patients with low CD4+ lymphocytes counts, suggesting a non T cell-mediated effect 16. Fifteen patients with disseminated MAC and other nontuberculous mycobacteria infections were treated subcutaneously during one year or more, 13 of them improved and 7 had even apparent disease eradication 20. Other groups of HIV negative immunocompromised patients infected with M. avium received clinical and bacteriological benefits of this treatment as well 1718. Short courses of the cytokine administered by alveolar route seemed to be effective in pulmonary MAC, since their use in one patient 19. However, it has been recently reported that a randomized trial testing this option was stopped early due to lack of efficacy 44. Since different routes of administration (aerosols, subcutaneous, intramuscular) have been used, this aspect should be further investigated.

Oxidative stress indicates an excessive production of oxygen reactive species (ROS), together with an inadequate antioxidant defense 45. Scarce information exists about pulmonary infections including M. avium in this sense. The redox unbalance was evidenced through oxidative damage markers increments, as shown by the comparison with healthy control reference values. AOPP are considered as the major free radicals-dependant protein damage indicator 46. As result of the exposition to radicals, proteins could suffer modifications, fragmentation, aggregation, changes on the absorption, decrease or loss of their biological function or proteolysis 47. After treatment, a significant reduction was detected in the IFN group. Their decrease or return to normal values can be secondary to infection eradication. In the same group, a SOD reduction was also observed. Although SOD plays a role in the defense against toxic oxygen metabolites, in most of the cases its presence indicates tissue damage and participates in reactions that drive to formation of hydrogen peroxide and iron, producers in turn of new toxic molecules 48. Therefore, their decrease can be taken as another favorable presage.

Genetic modulation data obtained are consistent with the clinical results. A significant increment in TGF beta was obtained in the placebo group. This factor counteracts T-cell proliferation and the production of IFN gamma, TNF alpha and pro-inflammatory chemokines 49, in addition to its pro-fibrotic activity 33, and thus helps Mycobacteria to avoid immune reactions and survive. MCP-1 participates in leukocyte recruitment toward granuloma 50. It also inhibits IL-12 production favoring infection 51. Therefore, MCP-1 decrease in the IFN group could be associated to pathogen control leading to granuloma resolution. The same interpretation is possible for the TNF alpha increase that occurred in the placebo group but not in the IFN treated patients. Although IFN gamma has been considered as a pro-inflammatory molecule, more and more evidences of anti-inflammatory actions exist, which supposes a dual effect 52.

The results obtained may justify the rationality to use IFN gamma as adjuvant to antimycobacterial drugs in patients with pulmonary atypical Mycobacteriosis, mainly in MAC infection, which is the most frequent. Their combination could reduce treatment duration, toxicities and possible relapses. In some cases it could prevent recessional surgery. Further, more extensive, controlled clinical trials are encouraged to confirm this assessment.

AOPP: Advanced oxidation protein products; BMI: Body mass index; BMRC: British Medical Research Council; CCR: Chemokine (CC motif) receptor; MAC:Mycobacterium avium complex; MHC: Major histocompatibility complex; GAPDH: Glyceraldehyde-3-phosphate dehydrogenase; GSR: Globular sedimentation rate; HIV: Human Immunodeficiency Virus; IFN: Interferon; IL: Interleukin; MDA: Malonyldialdehyde; MCP: Monocyte chemoattractant protein; PBMN: Peripheral blood mononuclear; PLPP: Potential lipid peroxidation products; ROS: Oxygen reactive species; SOD: Superoxide dismutase; STAT: Signal transducer and activator of transcription; TGF: Transforming growth factor; TNF: Tumor necrosis factor; T-bet: Transcription factor associated to Th1 lymphocytes; THP: Total hydroperoxides.

Authors IGG, CMVS, TRG, IBR, LGM and PALS are employees of the Center for Biological Research, which is part of the Center for Genetic Engineering and Biotechnology, Havana network, where human recombinant IFN gamma is produced. The rest of the authors have no competing interests at all.

MTMV conceived the study and was the main medical investigator. IGG participated in the study design, coordination and monitoring, and wrote the manuscript draft. YSH, MVQ, GJM, and NFO took care of patient recruitment, management, and follow-up. CMVS participated in the study design and results analyses. TIRG and IBR carried out molecular biology studies. RBSO did the radiological evaluations and CRA made the bacteriological determinations. LGM contributed as study monitor. GMS developed the oxidative stress evaluations. PALS took part in the design, results analyses and manuscript writing. All authors read and approved the final manuscript.

The other members of the MACGAM Study Group are: Roberto Suárez-Méndez, Dalia Carbonell-Freire (†), Nancy Silva-Sánchez, Delfina Machado-Molina, Eloína Turró-Soto, Helia I Herrera-García, and Madelyn Martínez-Soret from the "Benéfico Jurídico" Hospital, Havana, Mildrey Iglesias, "Amalia Simoni" Hospital, Camagüey, Elizeth García-Iglesias and Orlando C San Jorge-Mesa, Center for Biological Research, Havana.