American Journal of Respiratory and Critical Care Medicine

Despite the success of lung transplantation, infection is one of the leading causes of morbidity and mortality. Mycobacterial infections have been reported rarely, with the majority due to Mycobacterium tuberculosis. Our aim was to assess the incidence, etiology, and clinical outcome of mycobacterial infection after lung transplantation; to do so, we have studied retrospectively all lung and heart– lung transplants performed over a 12-yr period between November 1986 and June 1998 (n = 261). Twenty-three patients (9%) (M:F, 11:12) were diagnosed with mycobacterial infections in 25 sites, including n = 19, pulmonary (M. avium complex [n = 13], M. tuberculosis [n = 2], M. abscessus [n = 2], M. asiaticum [n = 1], and M. kansasii [n = 1]) and n = 6 extrapulmonary (M. haemophilum [n = 5] and M. abscessus [n = 1]) infections. Time to diagnosis from transplantation was 677 ± 735 d (range, 2– 3,086 d). Three episodes of transient colonization with M. avium were not treated; the remaining (22 of 25, 88%) were treated. Initial baseline therapy for nontuberculous mycobacteria included clarithromycin, rifampicin, ciprofloxacin, and/or ethambutol. All cutaneous lesions resolved completely, while clinical and graft function improved in 11 of 16 (69%) and 8 of 16 (50%) of patients treated, respectively. Seventeen of 23 patients (72%) survived at a follow-up of 1,658 ± 759 d (range, 522–3,285 d). Complications, predominantly due to rifampicin, included gastrointestinal intolerance and an increased tendency for rejection. There were no deaths attributable to mycobacterial disease or therapy. We conclude that mycobacterial infection, particularly due to nontuberculous mycobacteria, is relatively common after lung transplantation and may be an unrecognized cause of graft dysfunction. Early treatment of cutaneous lesions is associated with excellent control; however, graft dysfunction may be permanent. Although drug toxicity and interactions with immunosuppressive agents were not infrequent, the majority of these infections can be managed successfully. Malouf MA and Glanville AR. The spectrum of mycobacterial infection after lung transplantation.

Lung transplantation is now an accepted and successful method of treatment for end-stage respiratory failure due to primary pulmonary parenchymal disease or pulmonary hypertension. Infection is one of the leading causes of early and late morbidity and mortality in this group. The spectrum of pathogens includes bacteria, viruses, fungi, and protozoa (1, 2). Mycobacterial infections have been reported rarely, with most reported cases due to Mycobacterium tuberculosis (3). The treatment of mycobacterial infection may be difficult in the normal host, but treatment in the immunosuppressed population is associated with even more therapeutic limitations due to drug interactions with immunosuppressive agents and enhanced toxicity. We have experienced a significant number of mycobacterial infections over a 12-yr period in our lung and heart–lung transplant group. The majority of these infections were due to nontuberculous mycobacteria. We have been able to manage these infections successfully, leading to a marked improvement in the clinical status of the majority of patients, with no deaths attributable to mycobacterial disease or its therapy. This article, which is the largest series dealing with this potentially life-threatening disease (4) in lung transplant patients published to date, focuses on the diversity of etiological agents, their differing modes of clinical presentation, and the diagnostic and management problems associated with specific clinical outcomes in this group.

We reviewed all heart–lung and lung transplant patients (n = 261) transplanted between November 1986 and June 1998. Of these, 98 were single lung, 100 were bilateral lung, and 63 were heart–lung transplants. Twenty-three patients (9.0%) had 25 documented mycobacterial infections, with pulmonary isolates (n = 19) occurring more frequently than extrapulmonary disease (n = 6).

During the period under review, the methods used in the growth and identification of mycobacteria underwent modification. Before 1994 all specimens were inoculated onto two separate Lowenstein– Jensen (LJ) slants, one with added pyruvate and the other with glycerol. Nonsterile specimens were also decontaminated with alkali (NaOH) to remove organisms that would interfere with the isolation of mycobacteria. These were incubated at 37° C, except for skin and lymph nodes which were set up at 30° C. Chocolate agar slants were also inoculated and incubated at 30° C to facilitate the growth of Mycobacterium haemophilum. Cultures were reviewed weekly for 6 wk before being discarded. In mid-1994, BACTEC, an automated radiometric broth containing Middlebrook 7 H12 broth (Becton Dickinson, Franklin Lakes, NJ) was introduced, replacing the Lowenstein–Jensen with glycerol culture method. The BACTEC broth is suitable for mycobacterial cultures of virtually all specimen types including sputum, bronchoalveolar lavage, cerebrospinal fluid (CSF), biopsy tissues, and body fluids. The BACTEC broth was incubated for 6 wk and examined three times during the first week, followed by twice in the second week and weekly thereafter. All positive specimens were confirmed as acid-fast positive using the Ziehl–Neelsen stain and subcultured onto LJ slants. Subsequently, all culture-positive specimens were sent to a mycobacterial reference laboratory, the Institute of Clinical Pathology and Medical Research (ICPMR) Westmead Hospital, Westmead, Sydney for full identification. Positive BACTEC cultures were tested in the interim at St. Vincent's Hospital, using the NAP test (p-nitro-α-acetylamino-β-hydroxypropiophenone) over a period of 5 d, a biochemical test inhibiting only M. tuberculosis.

At the reference laboratory, identification was performed using conventional biochemical tests. Growth of M. tuberculosis and M. avium took approximately 3–6 wk and sensitivities took another 1–4 wk. Interim diagnosis of M. tuberculosis complex was made using DNA hybridization probes (Accuprobe gene probe; Accuprobe, Gen-Probe, Inc., San Diego, CA). Detection was obtained by chemiluminescence. This technology was also used to identify occasional strains of M. kansasii, M. avium, M. intracellulare, and M. gordonae. For noncultivatable organisms, DNA amplification by polymerase chain reaction (PCR) was performed using primers based on the 16S ribosomal RNA. Sequencing of PCR product was performed to confirm the presence of M. haemophilum. PCR methodology for the diagnosis of M. tuberculosis and M. avium complex was introduced at St. Vincent's from 1994, using Amplicor (Roche Molecular Systems, Branchburg, NJ). Results from the Accuprobe identification test were interpreted in conjunction with other laboratory and clinical data.

DNA probes were applied only to cultures growing mycobacteria whereas PCR could be used on most processed specimens (except feces). Sensitivity test results were available after 1 wk for M. tuberculosis; however, it was not usual for sensitivities to be performed on nontuberculous mycobacteria species unless known to be clinically significant or specifically requested. Criteria for diagnosis of mycobacteria disease were based on diagnosis and treatment of disease caused by nontuberculous mycobacteria (5), which included constitutional symptoms including cough, sputum, fever, weight loss, and dyspnea; radiographic changes such as pulmonary infiltrates with or without nodules, cavitation; bacteriological criteria (three positive sputum cultures with one acid-fast bacilli [AFB] smear negative, two positive sputum cultures with one AFB smear positive, or a single bronchial washing, with 1+ or greater growth); and positive tissue biopsy showing mycobacterial histopathologic features. This formed the basis of our judgment that the mycobacterial species could possibly be a significant contributor to the worsened clinical status of a patient and require therapy.

Chemotherapy was usually initiated on an outpatient basis, with empirical antibiotics employed to cover a broad spectrum of possible mycobacterial infections until full identification was available. Rifampicin (450 mg daily), clarithromycin (500 mg twice daily), ethambutol (15 mg/kg or less daily), and ciprofloxacin (500 mg twice daily), were the usual starting doses. The initiation of therapy was not delayed to wait for the full identification of mycobacterial species or their complete sensitivities. Alterations of therapy were common and varied according to species identification, evidence of clinical improvement, or patient tolerance of the medications.

Statistical Analysis

The data are expressed as means ± SD and range, unless otherwise stated.

Patient Population

Emphysema was the most common underlying disorder leading to lung transplantation in this case series. Only patients transplanted before July 1995 received induction therapy (anti-lymphocytic globulin, ALG) varying from 1 to 7 d. All patients received a standard maintenance triple immunosuppression regimen of cyclosporin (3–5 mg/kg), azathioprine (1–3 mg/kg), and prednisolone (0.2–0.5 mg/kg). All patients had a Mantoux test performed as part of their pretransplant workup.

Clinical presentation. For clinical correlation, please refer to Tables 1 and 2, where all cases are listed and numbered accordingly.

  1. 1. Patient 1 had a surveillance bronchoscopy with the diagnosis of asymptomatic cytomegalovirus (CMV) pneumonitis on transbronchial lung biopsy. The patient was clinically well.

  2. 2. Patient 2 was diagnosed 2 d after transplantation from examination of the explanted lung. Fifteen months later the same patient, whose transplant had been complicated by the development of an anastomotic stricture that required balloon dilatation and stenting, presented with cough, sputum, and worsening lung function.

  3. 3. Patients 3 and 5 were clinically well at surveillance bronchoscopy.

  4. 4. Patients 4, 7, and 18 had been unwell for several weeks, with purulent sputum, fevers, worsening pulmonary function, and infiltrates on chest X-ray.

  5. 5. Patients 6, 8, and 16 had a rapidly progressive process resulting in severe graft dysfunction after a presumed viral illness and underwent transbronchial lung biopsies with lavage.

  6. 6. Patient 9 had a background of gastric reflux and recurrent aspiration secondary to a postoperative vagal nerve palsy with chronic cough and sputum on a background of slowly worsening graft function.

  7. 7. Patient 10 developed anastomotic problems as a result of ischemic bronchitis, which required stenting; occurring on a background of recurrent multiresistant Staphylococcus aureus, aspergillus and cytomegalovirus infections. Graft function continued to deteriorate despite appropriate antibiotic therapy.

  8. 8. Patient 11 had cough and purulent sputum, associated with increased fatigue; sputum culture identified S. aureus as well as acid-fast bacilli.

  9. 9. Patient 12 had bronchopneumonia at the right lung base resulting from Streptococcus pneumoniae. Despite almost complete resolution on treatment, the patient then developed increasing infiltrates involving upper zones bilaterally.

  10. 10. Patient 13 had required stenting for anastomotic bronchomalacia. Bronchoscopy revealed multiple species of bacteria for which the patient received appropriate antibiotics.

  11. 11. Patient 14 had been treated recently for cytomegalovirus infection. Graft function improved initially, then deteriorated with the development of purulent sputum, wheeze, and increasing dyspnea. Bronchoalveolar lavage and sputum cytology revealed significant quantities of Aspergillus, which was treated with liposomal amphotericin.

  12. 12. Patient 15 had chronic sputum production thought to be due to basal bronchiectasis; however, his lung function continued to fall despite appropriate antipseudomonal antibiotics.

  13. 13. Patient 17 presented with rapid onset of a large, right-sided, bloody pleural effusion.

  14. 14. Patient 19 had established, severe airflow limitation and developed multiple painful lesions on his left lower limb.

  15. 15. Patients 20, 21, and 23 were otherwise well when they developed multiple subcutaneous lesions on their limbs. The nodules were painful, erythematous, or violaceous.

  16. 16. Patient 22 developed lesions in the area of a recently removed central line.


PatientAgeSexDiagnosisTransplant TypeBOS GradeSite

Definition of abbreviations: A = α1-antitrypsin deficiency; B = bronchiectasis; BOS = bronchiolitis obliterans syndrome; BSL = bilateral single lung; CF = cystic fibrosis; CHD = congenital heart disease; Em = emphysema; HL = heart–lung; PPH = primary pulmonary hypertension; SL = single lung.


Graft DysfunctionChest X-RayPositive SpecimensAFB Smear* Specimens Analyzed
 1YesNilNAD1 BAL1+ 71
 2   iNo+Nil2 × 1 cm peripheral lesionsExplanted lungN/AExplanted lung
  iiYes++++Mid- and basal infiltrates1 BAL/1 sputum2+126
 3NoNilNAD1 BAL2+ 72
 4Yes++++Apical pleural thickening2 BAL3+131
 5NoNilNAD1 BAL2+ 53
 6Yes++++Basal infiltrates1 BAL2+ 93
 7Yes++++Basal infiltrates1 BAL2+2+ 72
 8Yes++++Apical and mid-zone infiltrates1 Sputum1+ 71
 9Yes++++NAD1 BAL2+ 83
10Yes+++++Mid- and basal infiltrates2 BAL2+ 91
11Yes+++++NAD2 Sputa2+3+ 35
12Yes+++Bilateral apical infiltrates1 BAL3+ 81
13Yes++++Consolidation in midzones1 BAL3+ 42
14Yes+++++Bibasal infiltrates1 BAL/1 sputum2+3+ 22
15Yes++++++++Chronic bibasal bronchiectasis2 Sputa2+ 62
16No++NAD1 BAL/3 sputa2+2+ 53
17Yes++++Large right pleural effusionPleural fluid ×21+2+Pleural fluid × 2
18Yes++++++Infiltrates in transplanted lung2 Sputa1+2+135

Definition of abbreviations: AFB = acid-fast bacilli; BAL = bronchoalveolar lavage; N/A = not applicable; NAD = no abnormality detected.

*Semiquantitative assessment (range − to +++).

Mycobacterial cultures were performed on all patients from July 1987 onward as clinically indicated. All of the isolates occurred between December 1991 and March 1998, with a peak incidence of five cases in 1995. Time to detection of mycobacterial infection postoperatively was 568 ± 735 d (range, 2– 1,359 d). Time to complete identification was a further 52 ± 39 d (range, 13–91 d). The majority of patients who had mycobacterial infection isolated on bronchoalveolar lavage had evidence of graft dysfunction. Duration of follow-up was 1,658 ± 759 d (range, 522–3,285 d).

All specimens obtained were sent for microbiological analysis except those of Patient 2, a female with emphysema, who underwent a single lung transplant. A nodule had been present in the left upper lobe for at least 2 yr and was unchanged over this time. There were no known risk factors for mycobacterial infection and her Mantoux test was negative pretransplant. The explanted lung was placed in formalin for routine examination and, on review, the lesion had evidence of caseating granulomas staining positively with auramine stains. The specimen could not be sent for identification by microbiology methods and treatment was instituted for presumed M. tuberculosis. Three patients (cases 3, 5, and 13) had M. avium isolated from bronchoalveolar lavage on one occasion only and were clinically well; consequently they did not receive therapy.

Patient 21 developed multiple nodules on the ventrolateral aspect of her right forearm. Biopsies of a nodule showed multiple AFBs on Ziehl–Neelsen stain. Despite repeated biopsies, no growth of any organism occurred. Mycobacterium haemophilum was diagnosed when PCR using DNA hybridization was performed on the specimens 6 mo later.

These lesions did not respond adequately to the initial therapy of rifampicin, clarithromycin, and ciprofloxacin after 6 wk (Table 3). Consequently rifabutin (450 mg daily), a spiropiperidyl derivative of rifampicin that is known to be more effective against nontuberculous mycobacteria, was substituted for rifampicin. After 6 wk the patient complained of a blurred, painful, red left eye. A diagnosis of uveitis caused by rifabutin was made and treatment with prednisolone and homoatropine eye drops resulted in a rapid improvement in vision. The dosage of rifabutin was then reduced to 300 mg daily and the drug continued until the right eye also developed uveitis 2 mo later. Rifabutin was then ceased and the patient recommenced on rifampicin (300 mg daily) without further complications. It should be noted that the patient was also receiving triazole– itraconazole (400 mg daily), which can potentiate the toxicity of rifabutin.

Table 3. THERAPY

PatientPostoperative Days to DiagnosisMycobacteriumSourceTherapyDurationStatus/Days PostoperativeResponse to Therapy
 1  90 tuberculosis BALH, R, E, P, Pi 9 moAlive/2,358Excellent clinical response
  i   2 tuberculosis Lung explantedH, R, E, P, Pi 6 moAlive/1,332Asymptomatic clinical improvement,  minimal change in graft function
  ii 456 avium BALR, C, Ci27 moAlive/1,332Clinical improvement, minimal change  in graft function
 3 131 avium BALNilN/AAlive/1,661No treatment
 4 197 avium * BAL3/96 R, C, Ci, 3/96 + E,  7/96 − E, 2/97 − Ci,  3/97 − R28 moAlive/1,050Clinical improvement only
 5  51 avium BALNilN/ADead/522No treatment
 6 444MACBALR, C, Ci, Cl 2 moDead/593No graft or clinical improvement
 71,304MAC BALR, C, Ci, 6/96 + Cl,  1/97 − Cl, C34 moAlive/2,338No graft improvement
 81,211MACSputumR, C, Ci, − R, 7/9522 moDead/1,890Minimal clinical improvement
 9 152 avium BALR, E, C10 moDead/951No graft or clinical improvement
10 274 avium SputumR, E, C 6 moAlive/999No graft or clinical improvement
113,272 abscessus SputumRi, C, Ci 3 moAlive/3,363Sustained minimal graft improvement  significant clinical improvement
121,580 avium BALR, E, C, Ci 9 moAlive/2,035No graft improvement  significant clinical improvement
131,340 avium BALNilN/AAlive/2,037No treatment
14 360MACBALCl, R, Et, Ci 2 wkAlive/380No graft or clinical improvement
15 399 intracellulare SputumR, E, C12 mo Alive/792Excellent clinical and graft improvement
16  50 abscessus BALH, R, E, P, Pi  C 6 mo 54 mo Alive/1,874Excellent clinical and graft improvement
17 609 kansasii Pleural fluidH, R, E, C, Ci14 moAlive/1,470Excellent clinical and graft improvement
181,209 asiaticum Sputum4/94 + R, I, E, P, Pi44 mo Alive/2,672Excellent clinical and graft response initially
19 644 haemophilum LegH, R, Cl, P, Pi, 6/93 − H,  Cl, P, Pi, 6/93 + D, Ci,  10/93 + M − D, Ci, R42 moDead/2,194Complete resolution
20 324 haemophilum LegR, E, C17 mo Alive/1,114Complete resolution
 i 188 haemophilum * ForearmR, C, Ci, & 10/95 − Ci D31 mo Dead/1,140Complete resolution
 ii 310 haemophilum * Leg11/95 + Ri, − D, + E,  2/96 + C, 3/96 − E,  4/96 + E, 2/96 − Ri,  R, I, EComplete resolution
221,352 haemophilum Chest wallR, C, Ci, D18 moAlive/2,445Complete resolution
231,085 abscessusHandR, C, Ci33 moAlive/2,010Complete resolution

Definition of abbreviations: A = α1 antitrypsin deficiency; B = bronchiectasis; BAL = bronchoalveolar lavage; BOS = bronchiolitis obliterans syndrome; BSL = bilateral single lung transplant; C = clarithromycin; Ci = ciprofloxacin; CHD = congenital heart disease; CF = cystic fibrosis; Cl = clofazamine; D = doxycycline; E = ethambutol; Em = emphysema; H = isoniazide; HL = heart–lung transplant; M = minocycline; MAC = Mycobacterium avium complex; P = pyrazinamide; Pi = pyridoxine; PPH = primary pulmonary hypertension; R = rifampicin; Ri = rifabutin; SL = single lung transplant.

*DNA probe.


Current therapy.

Sixteen patients with pulmonary involvement received treatment. There was a marked and continued improvement in clinical signs and graft function in six (38%) (Figure 1), whereas in three (19%) graft function remained stable. However, despite a moderate improvement associated with stabilization for a period of months, three cases (19%) subsequently developed a slow decline in graft function over time. Six (38%) did not derive any sustained benefit from therapy, of whom Patients 6 and 14 received therapy for 8 wk or less because of drug intolerance.

The majority of side effects were due to intolerance to rifampicin. The use of clarithromycin and ciprofloxacin was also complicated by nausea and vomiting. This was overcome by introducing the drugs gradually and using a reduced dose. In spite of this, three patients were unable to tolerate rifampicin and it was discontinued. Those patients receiving clofazamine developed increased pigmentation of their skin. Other drug interactions include digoxin toxicity induced by renal impairment because of ciprofloxacin, possibly potentiated by itraconazole. The ethambutol dose was 15 mg/kg or less, aiming to avoid the development of optic neuritis. All patients taking ethambutol were evaluated and monitored by the ophthalmology service. Despite this, Patient 4 complained of blurred vision after 18 wk of ethambutol at 12 mg/kg. Ophthalmological review documented visual field loss and the drug was ceased with minimal residual field deficits.

We compared the incidence of rejection on the basis of events per 100 patient days, in those patients who developed mycobacterial infections versus the remainder of the lung transplant population. There was no significant difference between the linearized rates of rejection (Cox–Mantel) for the two groups to 36 mo. Six patients (30%) developed rejection during therapy. There was a total of nine biopsy-proven episodes of rejection, all involving Patients 2 and 21. The initial episodes were associated with subtherapeutic levels of cyclosporin. However, further episodes occurred despite adequate levels of cyclosporin and prednisolone and both of the patients were changed to tacrolimus (FK506) with no subsequent episodes of rejection.

Six patients in this series died; autopsies were performed on four of them. Patient 5, who was transplanted for cystic fibrosis, died of respiratory failure as a result of diffuse alveolar damage secondary to fulminant viral infection, and Patient 6 also died of respiratory failure secondary to upper lobe fibrosis and chondromalacia. Patient 19, with underlying α1-antitrypsin deficiency, died as a result of Pseudomonas bronchopneumonia. Patient 9, who was transplanted for primary pulmonary hypertension, also died of respiratory failure secondary to obliterative bronchiolitis and locally invasive aspergillosis. None of these patients had evidence of mycobacterial infection at the time of death, nor was it considered to have contributed to their deaths. Patients 8 and 21 did not have autopsies and died of respiratory failure as a result of obliterative bronchiolitis while waiting for retransplantation, and of overwhelming sepsis as a result of a lung abscess in the transplanted organ, respectively.

The pathogenicity of nontuberculous mycobacteria in this group is substantiated by the development of new pulmonary symptoms such as cough and sputum associated with radiographic infiltrates and repetitive isolation of the organism from pulmonary specimens and/or graft dysfunction. Although it was once commonly assumed that nontuberculous mycobacteria are capable of colonizing the normal host without causing invasive disease, approximately 90% of patients with disease caused by M. kansasii and most patients with M. avium complex have evidence of cavitation on computerized tomography (CT) scan (6).

Repeat Mantoux skin testing was not performed as it was felt to be of little value in this population, particularly as the majority was anergic. We found that mycobacterial infection generally occurred late in the posttransplantation period, and this has been observed in other studies (7). Although the finding of granulomatous disease on transbronchial lung biopsy or open lung biopsy may confirm the diagnosis of mycobacterial infection in those patients with worsening graft function (5), we have not had a similar experience. Despite the routine performance of transbronchial lung biopsy, all but one of our diagnoses were made from bronchial washings, pleural fluid, and sputa.

Patients presenting with extrapulmonary mycobacterial infections localized to the skin and subcutaneous tissues had no evidence of acute graft dysfunction. In a series reviewing nontuberculous mycobacterial infections in kidney, heart, and liver transplants (7) the incidence of pulmonary involvement was only 28% and more than 50% of patients also exhibited skin, joint, or soft tissue disease. Our series revealed infections predominantly involving the transplanted organ and there were no documented examples where pulmonary involvement coexisted with extrapulmonary disease.

To explain the variable response of graft function to therapy, we believe that as the majority of mycobacterial infections were found late, other processes had been established producing significant nonreversible airway damage leading to graft failure. Our results suggest that mycobacterial infections are one of many factors that cause airway damage and this, coupled with concurrent opportunistic infections and mechanical factors such as native lung hyperinflation, results in loss of lung function. Consequently progressive and irreversible loss of graft function was probably inevitable for the 6 of 16 patients (38%) treated with no clinical response. Importantly, 10 of 16 (62%) stabilized or improved.

Patient 18, who developed uveitis, was receiving multiple antibiotics, in particular clarithromycin, which is well known to potentiate the development of uveitis (8). The use of a triazole may also have potentiated the effect of rifabutin against M. avium complex (MAC), but may also have increased the risk for uveitis (9).

Rifampicin is well known to increase the metabolism of cyclosporin via induction of the hepatic cytochrome P-450 pathway, leading to low cyclosporin levels, which may cause rejection (10). In one study it reduced the area under the concentration–time curve of cyclosporin by nearly 60% and shortened its half-life from 4.1 to 1.8 h (11). Therefore close follow-up with frequent determination of cyclosporin levels is needed. Rifampicin administration also shortens the half-life of prednisolone (12). Subsequently the dose of cyclosporin was increased by a factor of three to four times and prednisolone doses were doubled. Rifampicin had a similar impact on tacrolimus. It was not necessary to cease antimycobacterial chemotherapy, particularly rifampicin, because of difficulties in maintaining adequate cyclosporin levels, which have been reported previously (3). A significant proportion of these patients was also receiving antifungals for coexistent fungal infections. Despite the fact that triazole antifungal agents (itraconazole and fluconazole) usually reduce cyclosporin requirements by 50–80% (13), this was not observed in those patients taking rifampicin. Rejection occurred when cyclosporin levels became subtherapeutic, compounded by lower bioavailability of prednisolone as a result of enzyme activity induced by rifampicin.

Mycobacterium tuberculosis was uncommon and occurred only in two patients, both of whom were managed successfully according to American Thoracic Society guidelines (14). Unlike M. tuberculosis, person-to-person transmission of nontuberculous mycobacteria is thought not to occur. However, the clustering of M. haemophilum cases (15) raises the possibility of a point source within the hospital. The spread of mycobacterial infection from a contact point has been raised in a series dealing with a contaminated water source in a renal dialysis unit (16). Nontuberculous mycobacteria are ubiquitous in the environment and the environment is the most likely source of infection. The donor organ may also be a source of infection which may have been of importance in Patient 1, who developed M. tuberculosis within 3 mo of being transplanted. Despite regular cultures for mycobacterial infections in this population, a diagnosis was not made before December 1991. The total number of lung transplants, however, was small, with only 25 heart–lung and 17 single lung transplants performed to January 1992, with a significantly higher early mortality rate. Bilateral lung transplants were not undertaken until June 1992. In addition, improved methods of diagnosis, including the introduction of PCR and DNA probes may have increased diagnostic yield.

It has now become clear that a macrolide antibiotic such as clarithromycin (17, 18) should be included in almost all drug regimens, particularly for MAC, and that rifabutin is superior to rifampicin in the treatment of MAC (19).

We have achieved successful outcomes in the majority of these patients. We selectively chose to observe and not treat those patients from whom we had obtained a single isolate of M. avium, providing mycobacteria were not found on follow-up bronchoscopic surveillance. These patients had no evidence of graft dysfunction, and were otherwise clinically well. However, in certain cases mycobacteria occurred in association with other infections in patients with severe graft dysfunction. The issue of whether mycobacteria may potentiate damage to the graft by other bacteria, viruses, or fungi is unanswered. However, as isolation of mycobacterial disease tended to occur late in our transplantation population, it is feasible that late airway damage may facilitate colonization with mycobacteria.

In summary, we report the largest series of mycobacterial infections after lung transplantation, emphasizing potential major problems. These include drug intolerance and interactions, particularly with immunosuppressive agents. Cutaneous lesions responded well to therapy; however, loss of pulmonary function was often permanent, perhaps reflecting other factors influencing graft survival. The use of rifampicin increased the risk of allograft rejection due to the development of subtherapeutic cyclosporin and prednisolone levels. Our experience demonstrates that mycobacterial infection is a common and potentially unrecognized problem in the lung transplant recipient, but once diagnosed it may be controlled if not cured. Individual patients may achieve outstanding clinical responses (Figure 1).

The authors appreciate the advice given by Dr. Thomas Gottlieb (Department of Microbiology Concord Hospital, Sydney), the clinical care given by other members of the Transplant Unit, and the support of the Division of Microbiology (St. Vincent's Hospital and ICPMR, Westmead Hospital).

1. Brooks R. G., Hofflin J. M., Jamieson S. W., Stinson E. B., Remington J. S.Infectious complications in heart-lung transplant recipients. Am. J. Med.791985412422
2. Dummer J. S., Montero C. G., Griffith B. P., Hardesty R. L., Paradis I. L., Ho M.Infections in heart-lung transplant recipients. Transplantation411986725729
3. Dromer C., Nasef S., Velly J. F., Martigne C., Couraud L.Tuberculosis in transplanted lungs. J. Heart Lung Transplant.1261993924927
4. Miller R. A., Louis L. A., Kline J. N., Geist L. J.Mycobacterium tuberculosis in lung transplant recipients. Am. J. Respir. Crit. Care Med.1521995374376
5. American Thoracic SocietyDiagnosis and treatment of disease caused by nontuberculous mycobacteria. Am. J. Respir. Crit. Care Med.1561997S1S25
6. Ahn C. H., McLarty J. W., Ahn S. I., Hurst G. A.Diagnostic criteria for pulmonary disease caused by Mycobacterium kansasii and Mycobacterium intracellulare. Am. Rev. Respir. Dis.1051982683714
7. Patel R., Roberts G. D., Keating M. R., Paya C. V.Infections due to nontuberculous mycobacteria in kidney, heart and liver transplant recipients. Clin. Infect. Dis.191994263273
8. Griffith D. E., Brown B. A., Girard W. M., Wallace R. J.Adverse events associated with high dose rifabutin in macrolide containing regimens for the treatment of Mycobacterium avium complex lung disease. Clin. Infect. Dis.211995594598
9. Narang P. K., Trapnell C. B., Schoenfelder J. R., Lavelle J. P., Bianchine J. R.Fluconazole and enhanced effect of rifabutin prophylaxis [Letter]. N. Engl. J. Med.33011199413161317
10. Modry D. L., Stinson E. B., Oyer P. E., Jamieson S. W., Baldwin J. C., Shumway N. E.Acute rejection and massive cyclosprine requirements in heart transplant recipients treated with rifampicin. Transplantation3931993313314
11. Offerman G., Keller F., Molzahn M.Low cyclosporine A blood levels and acute graft rejection in a renal transplant recipient during rifampin treatment. Am. J. Nephrol.51985385387
12. Buffington G. A., Dominguez J. H., Piering W. F., Herbert L. A., Kauffman M., Lemann J.Interaction of rifampicin and glucocorticoids. J.A.M.A.23617195819581960
13. Kramer M. R., Merin G., Rudis E., Bar I., Nesher T., Bublil M., Milgalter E.Dose adjustment and cost of itraconazole prophylaxis in lung transplant recipients receiving cyclosporine and Tacrolimus (FK506). Transplant Proc.29199726572659
14. American Thoracic SocietyTreatment of tuberculosis and tuberculosis infection in adults and children. Am. J. Respir. Crit. Care Med.149199413591374
15. Saubolle M. A., Kiehn T. E., White M. H., Rudinsky M. F., Armstrong D.Mycobacterium haemophilum: microbiology and expanding clinical and geographic spectra of disease in humans. Clin. Microbiol. Rev.91996435447
16. Gouby A., Branger B., Oules R., Ramuz M.Two cases of Mycobacterium infection in a renal dialysis unit. J. Med. Microbiol.2519884649
17. Perronne C., Gikas A., Truffot-Pernot C., Grosset J., Vilde J. L., Pocidalo J. J.Activities of sparfloxacin, azithromycin, temafloxacin and rifabutin compared with that of clarithromycin against multiplication of Mycobacterium avium complex in human macrophages. Antimicrob. Agents Chemother.35199113561359
18. Fernandes P. B., Hardy D. J., McDaniel D., Hanson C. W., Swanson R. N.In vitro and in vivo activities of clarithromycin against Mycobacterium avium. Antimicrob. Agents Chemother.33198915311534
19. Brogden R. N., Fitton A.Rifabutin, a review of its antimicrobial activity, pharmacokinetic properties and therapeutic efficacy. Drugs47619949831009
Correspondence and requests for reprints should be addressed to Dr. M. A. Malouf, Cameron Wing 15, St. Vincent's Hospital, Victoria Street, Darlinghurst NSW 2010, Australia. E-mail:


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record