INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Mycobacterium avium complex (MAC) pulmonary disease is the most common nontuberculous mycobacterial infection in the United States and Japan, and is increasing in incidence (1, 2). Treatment of MAC pulmonary disease has been difficult, because antituberculous drugs are not as effective as they are in treating pulmonary tuberculosis (3). Antituberculous regimens consisting of three to five drugs have long been used empirically, but rates of successful treatment were variable (sputum conversion rate of 25 to 90%) and there were frequent relapses (4).

Recently, Wallace and colleagues (11) and Dautzenberg and colleagues (12) reported that clarithromycin is clinically efficacious in the treatment of MAC pulmonary disease as a single agent, but that clarithromycin resistance developed in failed cases. For heightening the efficacy and preventing the development of clarithromycin resistance, Wallace and colleagues initiated clinical trials of clarithromycin-containing regimens including clarithromycin, ethambutol, rifabutin or rifampin, and initial streptomycin, and reported that the regimen achieved satisfactory results (13). The American Thoracic Society recommended a three-drug regimen including clarithromycin or azithromycin, rifampin or rifabutin, and ethambutol for pulmonary MAC disease in 1997 (14). However, reports that substantiate the clinical efficacy of the clarithromycin-containing regimen for pulmonary MAC disease are few.

We initiated a single-center, open, noncomparative prospective trial of a clarithromycin-containing regimen including clarithromycin, rifampin, ethambutol, and initial kanamycin with subsequent ofloxacin, for pulmonary MAC disease in November 1992. We included patients who received treatment for the first time and patients who had already been treated at other hospitals with regimens including clarithromycin.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Enrollment of Patients

Patients in whom MAC pulmonary disease was diagnosed at the Chest Disease Research Institute, Kyoto University, or referred patients who had already been given a diagnosis of MAC pulmonary disease at other hospitals were enrolled sequentially from November 1992 to November 1997. Past history was obtained by direct interview, and history of prior treatment was confirmed by communication with a prior attending doctor. The diagnostic criteria were the presence of two or more positive sputum cultures for MAC and abnormal chest radiograph findings consistent with MAC pulmonary disease, or presence of granuloma with a positive culture for MAC. Patients were required to have negative serologic findings for human immunodeficiency virus (HIV) type 1 and type 2 or the absence of obvious risk factors for that disease.

Inclusion criteria for this study included positive sputum cultures for MAC at the time of entry into the study. Thus, patients who could not expectorate sputum and were given the diagnosis only by bronchial washing and transbronchial lung biopsy were excluded. Other inclusion criteria included availability for long-term follow-up and aggravation of symptoms and/or chest radiograph findings. Patients were included even if they had received prior therapy with any drug regimen. We defined prior therapy as 6 mo or more of drug therapy. In vitro susceptibility tests of MAC isolates were not performed at study entry. Instead, initial isolates obtained at the entry were stored at 80° C and susceptibility tests were done in 1998. Patients were excluded if they had a history of allergy to macrolides, or their serum creatinine level, total bilirubin level or serum aminotransferase levels were more than twice the upper limits of the normal range. Patients who had difficulty in visiting the hospital regularly due to geographic distance or who had low activity of daily living because of serious associated disease were also excluded. Informed consent was obtained from each patient at study entry under protocols approved by the Ethical Committee of the Chest Disease Research Institute, Kyoto University.

Protocol

A four-drug regimen that consisted of clarithromycin 10 mg/kg of body weight, ethambutol 15 mg/kg, rifampin 10 mg/kg given daily by way of mouth, and kanamycin 20 mg/kg intramuscularly given 3 times a week for the initial 2 to 6 mo of therapy, and subsequent daily quinolone, was used as a standard regimen. As quinolone, ofloxacin 10 mg/kg orally was used from November 1992 to June 1994, and an equivalent dose (5 mg/kg) of levofloxacin (the L-isomer of ofloxacin) was used instead of ofloxacin from July 1994. The overall treatment period was 24 mo.

A modified four-drug regimen was begun when hearing loss and/ or visual disorder were pointed out at the pretreatment evaluation, or when there was a history of drug allergy to companion drugs, or when a patient refused the frequent visits to the outpatient clinic for intramuscular injection of kanamycin. Alternative companion drugs were chosen from enviomycin, isoniazid, and ethionamide. Rifabutin and clofazimine have not been approved for clinical use by the Ministry of Health and Welfare in Japan. The choice of companion drugs was left to the attendant doctor's discretion.

Patients were admitted to the Chest Disease Research Institute Hospital for the evaluation of baseline conditions and monitoring of compliance and adverse effects at the introduction of the therapy. Routine expectorated sputum was submitted for examination on three consecutive days at entry into the study. An early-morning specimen was collected and processed as previously described (15). Briefly, the sputum was examined with Ziehl-Neelsen staining. Specimens submitted for culture were digested and decontaminated by the sodium hydroxide method, and the samples were inoculated onto slants of 3% Ogawa egg medium (Japan BCG, Tokyo, Japan). BACTEC/broth media, which are more sensitive than the agar, were not used. Mycobacteria were identified and differentiated by growth characteristics, conventional biochemical tests, and the Accuprobe test (Gen-Probe, San Diego, CA). M. avium and M. intracellulare were identified by the Amplicor polymerase chain reaction (PCR) assay (Roche Diagnostic Systems, Inc., Branchburg, NJ). Patients were admitted for the first 2 to 8 wk and examined every 2 wk at our clinic for the first 2 mo, every 4 wk thereafter during the therapy, and at least every 3 mo during the follow-up after completion of the therapy. Sputum was submitted for examination at every visit. When a patient could not expectorate sputum at the outpatient clinic, a sputum sample pot was handed to the patient, who was requested to submit a sputum sample to our microbiology laboratory by or at the next visit. Computed tomography (CT) scans were performed initially in all the patients for evaluating the lesions including bronchiectasis and cavity, and underlying pulmonary conditions. A chest radiograph was taken initially and at least every 3 mo thereafter. Disease extent was evaluated on the basis of chest roentgenogram findings. Complete blood count, aminotransferases, total bilirubin and serum creatinine tests were performed at each visit during the therapy. Patients were advised to consult an otolaryngologist and an ophthalmologist initially, and as needed thereafter. Visual acuity test, perimetry, and red-green color discrimination test were performed routinely to check for ethambutol toxicity at least every 2 mo.

Assessment of Efficacy

Efficacy was evaluated on the basis of sputum conversion rate. Sputum conversion was defined as consecutive negative sputum cultures during a 3-mo period, with the time of conversion defined as the date of the first negative culture. When the patient's condition improved and excretion of sputum disappeared completely, sputum induction with sterilized 3% NaCl was performed. If the patient could not expectorate sputum even after the sputum induction, it was considered that sputum had converted to negative. Sputum relapse was defined as two consecutive positive cultures after sputum conversion.

Withdrawal from the Study

Patients were withdrawn from the study when any of the following occurred: a request for withdrawal from the study made by the patient, loss to follow-up, death, or adverse events requiring discontinuation of clarithromycin. When companion drugs could not be used because of adverse events, a modified four-drug regimen of clarithromycin plus three drugs was adopted. When patients could not tolerate a regimen consisting of four drugs, they were withdrawn from the protocol and the therapy was individualized.

Clarithromycin Susceptibility

The MAC isolates obtained at the entry were stored in 85% 7H9 medium:15% glycerol at 80° C. The isolates obtained from the patients who had finished 24 mo of treatment with the standard regimen or modified four-drug regimen were also stored. The in vitro activity of clarithromycin against stored isolates was determined by the broth microdilution method, using twofold drug dilutions in Middlebrook 7H9 broth as previously described (16). The 7H9 broth was supplemented with Bacto Middlebrook ADC enrichment (Difco Laboratories, Detroit, MI), but no Tween 80 or other surfactants were added. Minimal inhibitory concentrations (MICs) were determined at pH 7.4. When a strain did not grow well at pH 7.4, MIC was determined at pH 6.8. Isolates were judged as clarithromycin-susceptible if they had MICs of 8 µg/ml or less and resistant if they had MICs of more than 32 µg/ml. MICs of 16 and 32 µg/ml were considered intermediate.

Statistical Analysis

All results are presented as mean ± SD. A comparison of characteristics of the patients by sex (Table 1) was done by the unpaired Student's t test. A comparison between the patients who converted and those who failed to convert was performed to identify factors that significantly affected the treatment success (Table 2). The Fisher's exact test with Yates correction as needed was used for categorical variables, and the unpaired Student's t test was used for continuous variables. The time from initiation of therapy to sputum conversion to negativity was expressed in the form of a Kaplan-Meier curve stratified by history of previous treatment (Figure 1), and the probability of sputum clearance was compared using the generalized Wilcoxon test. A value of p < 0.05 was considered statistically significant. Multivariate analysis was not performed because of the small sample size (17).

View larger version (16K): [in this window] [in a new window]   Figure 1.   Time from initiation of therapy to sputum conversion to negativity stratified by history of prior therapy [Kaplan-Meier curves: pretreatment () versus pretreatment (+)].

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients/Disease

From November 1992 to November 1997, 56 patients were newly given a diagnosis of MAC pulmonary disease at our hospital, and 31 patients had been given the diagnosis at other hospitals and were referred to our hospital. Among the 87 patients, 41 patients were ineligible and the primary reasons for ineligibility were as follows: minimal symptoms and no signs of aggravation during the last 6 mo in 14 patients, given the diagnosis using fiberoptic bronchoscopy in 13, difficult to visit regularly (geographic distance, association of serious disease with low activity of daily living) in seven, noncompliance in three, hepatic/renal disorder in two, history of macrolide allergy in one, comorbidity of tuberculosis in one. The other 46 patients (52.9%) fulfilled the eligibility criteria and were enrolled in the protocol. Baseline characteristics of the eligible 46 patients are shown in Table 1. The mean body weight of our patients was 46.1 ± 8.8 kg (32 to 62 kg), and 9.1 ± 1.6 mg/ kg of clarithromycin (400 to 600 mg daily) was used on average. All the patients were Japanese with a mean age of 60.9 ± 11.5 yr. There were more than twice as many women as men, and there was no significant age difference between the sexes. Twenty-two patients had underlying conditions: seven had healed tuberculosis, three had bronchiectasis, two had diabetes mellitus, one had gastrectomy, one had healed tuberculosis and diabetes mellitus, one had healed tuberculosis and gastrectomy, one had chronic obstructive pulmonary disease, one had idiopathic pulmonary fibrosis, one had mild hepatitis, one had lung lobectomy due to previous MAC infection, one had Sjögren's syndrome and pulmonary fibrosis, one had rheumatoid arthritis and had undergone operation for uterine myoma, one had a left radical mastectomy. Men were significantly more likely than women to smoke and to have underlying conditions. Some patients drank alcohol on social occasions, but there was no alcohol abuse.

Therapy

Among the 46 enrolled patients, 29 patients were begun on the standard regimen. Among the 29 patients, three patients were dropped and five patients needed modification of companion drugs within 6 mo of therapy because of adverse events: hearing loss and skin eruption in one patient, hearing loss in one, impaired vision in one, skin eruption in one, and hepatitis in one. The other 17 of 46 patients were begun on a modified four-drug regimen because of preexisting conditions including presbyacusis in eight patients, presbyacusis and retinal detachment in one, contraction of visual field in one, and because seven patients refused frequent visits for kanamycin injection. Among the 17 patients begun on a modified four-drug regimen, the initial regimen had to be remodified in one patient within 6 mo of therapy, and four patients were dropped. In total, 39 patients received more than 6 mo of therapy: the standard regimen for 21 patients (45.7%) and a modified four-drug regimen (including a remodified regimen) for 18 patients (39.1%), and seven patients (15.2%) were dropped within 6 mo. The reasons for withdrawal were as follows: adverse event due to plural companion drugs in three patients, clarithromycin allergy in one, fear of adverse event in one, association of other disease in one, and unrelated death in one. The contents of regimens are listed in Table 3. Of the seven patients dropped from the protocol, two patients received a three-drug regimen, three patients received a two-drug regimen, and two patients refused further treatment for fear of adverse events.

Laboratory

Of the 39 patients who received more than 6 mo of therapy, 23 patients had nodules and bronchiectasis disease and 16 had upper lobe cavitary disease. Minimal disease was present in four (10.3%), moderately advanced disease in 26 (66.6%), and far advanced disease in nine (23.1%).

Among the strains isolated from the 39 patients, one strain did not grow well in the broth at pH 7.4, and species and MIC were determined at pH 6.8. Of the 39 pretreatment isolates, 26 strains were M. avium and 13 were M. intracellulare. Among them, 31 strains were susceptible, two were intermediate, and six were resistant (MIC > 32 µg/ml) to clarithromycin.

Sputum Conversion

Among 31 patients whose pretreatment isolates were susceptible to clarithromycin, 24 converted by 6 mo and two additional patients converted by 9 mo. Hence, five patients (16.1%) failed to convert in the end. Three of the five patients had extensive disease, and MAC strains obtained at the end of the regimen were clarithromycin-resistant in the three patients. The two other failed patients had moderate disease, and the posttreatment strains were susceptible to clarithromycin. The reason for the failure is unknown.

Microbiological relapse occurred in three patients and clarithromycin-resistant strains were obtained at the relapse in two of the three patients. Two patients relapsed 4 and 7 mo each after completion of a full course of therapy. The former patient was a 59-yr-old female who had a history of pretreatment. Ethambutol could not be administered because of adverse effects and the regimen was modified. She is under observation without therapy because there is no aggravation of symptoms or chest roentgenogram findings. The relapsed isolate was susceptible to clarithromycin. The latter was a 66-yr-old male who had received pretreatment for 3 yr. He had moderately advanced disease with huge cavities. He is under retreatment with a modified regimen because chest roentgenogram findings worsened with a microbiological relapse. The relapsed isolate was resistant to clarithromycin. The other relapse occurred at 8 mo after conversion during the therapy. The patient was a 69-yr-old female who had never been treated. Ethambutol and kanamycin could not be administered because of adverse effects and the regimen was modified. The pretreatment isolate was susceptible to clarithromycin, and the isolate obtained after relapse was resistant. She is still continuing therapy with a modified regimen. The three relapsed patients were adherent to the therapy and had no diarrhea/malabsorption disease. In total, resistance to clarithromycin newly developed in five patients: three of five patients who did not convert and two of three patients who relapsed after conversion.

Among six patients infected with resistant strains, one patient converted her sputa to negative but the other five failed to convert. All six patients were referred patients and had had clarithromycin-containing therapy for MAC pulmonary disease: one patient with clarithromycin plus quinolone for 4 mo, two with clarithromycin monotherapy for 6 and 12 mo each, one with rifampin plus clarithromycin for 6 mo, and the other two with histories of prior therapy for MAC disease for more than 10 yr with many kinds of regimens, including clarithromycin monotherapy or clarithromycin plus one companion drug. Among two patients whose pretreatment isolates showed intermediate susceptibility to clarithromycin, one patient failed to convert.

In total, 28 (71.8%) of 39 patients who received more than 6 mo of therapy converted their sputa to negative, and the timing of sputum conversion was 3.6 ± 1.9 mo, with a range of 2 to 9 mo. The conversion rate was significantly lower in patients who had had prior therapy (55.0% versus 89.5%, p < 0.05), who were infected with clarithromycin-resistant or intermediate strains other than those with susceptible strains (25.0% versus 83.9%, p < 0.01), or who were acid-fast bacilli (AFB) smear-positive at the entry (60.7% versus 100%, p < 0.05), as shown in Table 2. Age and sex of patients, species of pathogen (M. avium or M. intracellulare), type and extent of disease, and choice of kanamycin as a companion drug did not significantly affect the conversion rate. Figure 1 shows the time from initiation of therapy to sputum clearance stratified by history of prior therapy. Newly treated patients had a significantly higher probability of sputum clearance of MAC than retreated patients (p < 0.01).

Adverse Events

Clarithromycin was discontinued in three of 46 enrolled patients. Skin rash developed in one patient. Hearing loss which had been caused by streptomycin at prior therapy was aggravated in one patient who received treatment with clarithromycin plus rifampin, ethambutol, and levofloxacin, and she refused further therapy. Watery diarrhea continued in another patient who received treatment with clarithromycin plus rifampin, ethambutol, and kanamycin, and she also refused further therapy. It is not clear whether the adverse events were caused by clarithromycin in the latter two cases.

Adverse events resulting from companion drugs occurred in 10 of the 46 patients (21.7%). Six of them were able to continue a four-drug regimen with modification of companion drugs, but three patients withdrew because plural drugs caused adverse events, and one patient withdrew because of adverse events plus death due to progression of underlying idiopathic pulmonary fibrosis. Among companion drugs, ethambutol caused decrease of visual acuity in three patients, skin eruption in two, and contraction of visual field in one; kanamycin caused hearing loss in two and vestibular dysfunction in one; levofloxacin caused skin rash in two, and erosive gastritis in one; rifampin caused skin rash in one and gastrointestinal symptoms in one; and ethionamide caused hepatitis in one and symptomatic hypothyroidism in one.

Current Status

Of the 39 patients who received more than 6 mo of therapy, four patients discontinued after 6 to 24 mo of therapy, 21 patients completed the full course of therapy, and the other 14 patients are receiving therapy (mean 14.7 ± 4.5 mo, at the end of October 1998) as shown in Table 4. In total, 22 of the 39 patients (56.4%) remain culture-negative to date. Treatment was finished at 18 mo in one patient who converted sputum to negative by 6 mo of therapy with complete resolution of abnormal shadow on chest radiograph by 12 mo. The other 20 patients received 24 mo of therapy.

The primary reasons for withdrawal after 6 mo were unrelated death at 12 mo in one case, refusal of further therapy at 14 mo owing to fear of adverse events in one, loss to follow-up at 16 mo in one, and death secondary to rapid deterioration of MAC pulmonary disease at 12 mo in one.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have investigated the efficacy of a clarithromycin-containing four-drug regimen for MAC pulmonary disease in patients without acquired immunodeficiency syndrome (AIDS). So far, two groups have reported that clarithromycin is efficacious for pulmonary MAC disease. Dautzenberg and colleagues used approximately 30 mg/kg/d of clarithromycin for 45 patients either in monotherapy or with companion drugs (12). They reported that 36 of 39 patients (92%) converted their sputa to negative by 3 mo. However, the early failure or relapse rate was 18% and the investigators stated that monotherapy is not adequate and concomitant active drugs are needed to avoid the development of resistant mycobacteria (12). Wallace and colleagues first reported the efficacy of clarithromycin monotherapy (11) and then of clarithromycin-containing four-drug regimens for pulmonary MAC disease without AIDS (13). They reported intolerance of a high dose of clarithromycin (2,000 mg/d) among elderly patients in their earlier study (18) and used 1,000 mg of clarithromycin daily (11, 13). Although the mean body weight of patients was not specified, it would be approximately 60 kg judging from their prior report (18) and the dose of clarithromycin would be 15 to 20 mg/kg/d. They chose rifabutin or rifampin, ethambutol, and streptomycin as companion drugs of clarithromycin. Although our protocol was similar to their protocol, the overall conversion rate in our study was lower than theirs (72% compared with 92%). This may be a result of three major factors. The first is that we included patients whose isolates were resistant to clarithromycin, the second is a lower dose of clarithromycin, and the third is the choice of rifampin instead of rifabutin.

It is not clear whether clarithromycin should be withdrawn in cases with resistant strains. It is obvious that clarithromycin monotherapy is not effective for the treatment of infection with clarithromycin-resistant strains (11). However, the possibility of synergistic effects of multiple drug regimens on resistant strains has not been disproven. Moreover, there is a possibility of polyclonal MAC infections with mixtures of susceptible and resistant strains, especially in pulmonary disease with nodules and bronchiectasis (19). Thus, we included all the patients without performing susceptibility tests in advance. As a result, however, the conversion rate was significantly lower in patients infected with resistant strains (16.7%) and intermediate strains (50.0%) compared with those infected with susceptible strains (83.9%, p < 0.01), and these results suggest a lack of synergism or partial effects for polyclonal MAC infections. All the patients infected with clarithromycin-resistant strains had received prior therapy including clarithromycin monotherapy or clarithromycin plus one companion drug. These results suggest that clarithromycin plus at least two companion drugs should be used to avoid induction of clarithromycin resistance. If the clarithromycin resistance is once induced, salvage therapy is not likely to be effective. On the other hand, five of 31 patients infected with clarithromycin-susceptible strains failed to convert their sputa to negative. Three of the five failed patients had extensive disease and clarithromycin resistance was induced in the three patients. Clarithromycin resistance was induced also in two of three relapsed cases. Whether sputum conversion could have been achieved in such patients by increasing the dose of clarithromycin remains to be elucidated.

If we exclude the patients infected with clarithromycin-resistant and intermediate strains, the conversion rate in patients infected with susceptible strains was 26 of 31 (83.9%) as a whole. The rate is comparable to 10 of 13 (77%) in patients who received rifampin instead of rifabutin in the study by Wallace and colleagues (13). The comparability would suggest that a clarithromycin daily dose of 10 mg/kg versus 20 mg/kg might not have critically affected the conversion rate, on the condition that clarithromycin is used with three companion drugs. For determining the appropriate dose of clarithromycin, however, a controlled trial will be necessary.

Rifabutin is more active than rifampin against MAC in vitro (20). Moreover, rifampin is known to accelerate the rate of metabolism of clarithromycin and to reduce the clarithromycin concentration in serum more than rifabutin does (21). In the study by Wallace and colleagues (13), the rate of successful treatment was 22 of 26 (85%) in the rifabutin-containing regimen, and 10 of 13 (77%) in the rifampin regimen. Although the difference was not significant, rifabutin appears to be slightly superior to rifampin. The disadvantage of rifabutin is that it causes adverse events, including uveitis and leukopenia, which are infrequent with rifampin (13, 22). Because most patients with MAC pulmonary disease are elderly in Japan as in the United States, and because long-term therapy seems to be necessary for the disease, adverse events may lower the rate of treatment success by increasing the withdrawals or by lowering compliance. Furthermore, a recommendation to include rifampin or rifabutin in combination therapy may not be warranted. A controlled trial of rifampin-containing regimen versus rifabutin-containing regimen versus regimen without rifampin or rifabutin is necessary.

The roles of aminoglycosides and quinolones as companion drugs are also unclear. Wallace and colleagues recommended initial use of streptomycin in patients with extensive and/or cavitary disease (13). We have used kanamycin on the basis of previous data from in vitro and animal studies (23). There are, however, no clinical studies which have shown that kanamycin is superior to streptomycin or other aminoglycosides against MAC. In our study, the conversion rate was higher in patients who received initial kanamycin (77.8%) than in patients who did not receive it (58.3%), although the difference was not significant. However, elderly patients who had presbyacusis at the pretreatment evaluation were frequent and consequently, the use of kanamycin was often limited. Some patients refused frequent visits to the clinic for kanamycin injection over a long period. It would be reasonable to add aminoglycoside in patients who need hospitalization because of severe symptoms and/or signs.

Lastly, the most important and difficult question left unsolved is when to start therapy, as the American Thoracic Society recommendation also pointed out (14). In our study, the conversion rate was significantly higher in smear-negative and newly diagnosed cases. The result would suggest the earlier, the better. In the long prospect, however, there is no guarantee that relapse will never occur after completion of initial treatment, because underlying conditions, including preexisting lung diseases and unknown factors common to elderly women (24), cannot be removed, and because MAC is a ubiquitous pathogen. In patients who had prior therapy, conversion rate was significantly lower than in newly treated patients (55.0% versus 89.5%, p < 0.05). Hence, it is still not clear how to decide when to start therapy.

In conclusion, the clarithromycin-containing four-drug regimen was highly efficacious for the initial treatment of MAC pulmonary disease. However, the sputum conversion rate was unsatisfactory in patients who had prior therapy and those who were infected with clarithromycin-resistant strains. Development of new companion drugs that are as active as clarithromycin and that can be tolerated by elderly patients is needed to further increase the conversion rate of sputum and to avoid the induction of macrolide resistance.