Infection continues to be a major source of morbidity and mortality in solid-organ transplant recipients, who not only are predisposed to primary infections, but are also at risk for reactivating latent infections acquired before transplantation or transmitted via the donor organ (1).
Tuberculosis is unusual in these patients; however, the exact incidence has not been adequately defined as yet, and information regarding the specific problems encountered in the diagnosis of tuberculosis in solid-organ transplant recipients is scarce. It is also unclear whether there are differences in the manifestations of the disease between solid-organ transplant recipients and other immunocompromised or immunocompetent hosts. Furthermore, the optimal treatment for tuberculosis in these patients is not known, nor is the impact of tuberculosis and antituberculous drugs on the patient and graft survival.
The objectives of our study were to know the incidence of tuberculosis in solid-organ transplant recipients, the clinical consequences of the disease in this population, the risk factors for the development of tuberculosis, the interference of antituberculous drugs with immunosuppressive treatment, and the factors influencing mortality. We report 51 cases of tuberculosis from seven solid organ transplantation centers in Spain and review the literature on the subject.
PATIENTS AND METHODS
This retrospective study was carried out at seven university teaching hospitals in Madrid, Barcelona, Valencia, and Córdoba, Spain. The hospital records and outpatient follow-up were reviewed of all patients over 18 years old who received a kidney, liver, or heart transplant between 1980 and 1994, and developed active tuberculosis after transplantation. Specifically excluded were all patients in whom tuberculosis was diagnosed before transplantation. Cases were identified by reviewing the files of the kidney, liver, and heart transplantation programs of each institution. In addition, we reviewed all Mycobacterium-positive specimens obtained from solid-organ transplant recipients. Fifty-one patients were found to have posttransplantation tuberculosis during the period studied.
Data collection. Patient clinical charts were reviewed for demographic information, type of organ transplanted, degree of immunosuppression, purified protein derivative of tuberculin (PPD*) and delayed-type hypersensitivity skin testing, previous history of tuberculosis, chemoprophylaxis, time to development of tuberculosis (early: ≤1 year, late: >1 year), clinical presentation, laboratory data, radiographic and pathologic features, results of Mycobacterium cultures, time to reach diagnosis, sites of involvement of tuberculosis infection, methods for diagnosis, type and number of drugs for treatment, related toxicity, number of rejection episodes, microbiological and clinical outcome, and cause of death.
Furthermore, the English-language literature on tuberculosis infections from 1967 to 1994 was reviewed using MEDLINE. The key words tuberculosis, Mycobacterium tuberculosis and Mycobacterium infection were cross-referenced with the key word transplantation; secondary references were also reviewed. In addition, the articles containing the largest series of infectious complications after solid organ transplantation were reviewed. Cases were excluded from the literature review if tuberculosis was diagnosed before transplantation. Series with four or more cases of tuberculosis in organ-transplant recipients were tabulated and analyzed with respect to the incidence of tuberculosis, time from transplantation, type of tuberculosis, treatment-related toxicity, and outcome.
Case definition. The diagnosis of tuberculosis was considered certain if M tuberculosis was cultured from any clinical sample. Tuberculosis was considered probable in patients who had a clinical picture highly suggestive of tuberculosis that resolved with specific antituberculous treatment, and in whom acid-fast bacilli were seen on smear, with or without caseating granuloma on histopathology.
Disseminated tuberculosis was considered to be present when M tuberculosis was isolated from two or more noncontiguous organs or from blood, or when there was isolation of the organism from one organ along with demonstration of acid-fast bacilli or granuloma at a different site.
Microbiological identification. Samples were processed and organisms identified according to standard methods (2). Ziehl-Neelsen stain and the auramine O-rhodamine fluorescent method were used for staining. Samples were cultured in different media according to the laboratory and the year of diagnosis. Reasons for performing cultures to detect Mycobacterium included both clinical indications and surveillance for tuberculosis. Antimicrobial susceptibility testing of the M tuberculosis isolates was not routinely done.
Immunosuppression and immunological tests. Immunosuppressive regimens varied according to the hospital, the organ transplanted, and the year of transplantation. Triple therapy with cyclosporine (CsA), azathioprine, and steroids was used during the study period by the majority of centers. Doses of CsA were adjusted to obtain trough plasma levels of 200-400 ng/ml (as determined by radioimmunoassay) during the first month and 100-200 ng/ml thereafter. Rejection episodes were usually treated with boluses of steroids or antilymphocyte antibodies (antilymphocyte globulin, antithymocyte globulin, or OKT3) in the case of steroid-resistant rejection.
A positive response to tuberculin (PPD test) was defined as an induration of 5 mm or more in diameter, 48-72 hr after administration of 2 U of the RT-23 strain equivalent to 5 tuberculin units of PPD. Delayed-type hypersensitivity skin testing was performed using a multiple-punction skin test device (Multitest CMI; Rhone Poulenc Farma, SAE, Madrid, Spain) at the same time tuberculin was injected. Anergy was defined as the absence of induration ≥2 mm in diameter in response to any of the seven skin-test antigens at 48 hr. The cutoff point of 2 mm was chosen according to the manufacturer's recommendations. Since 1985, pretransplantation HIV screening has been performed routinely in all centers.
Follow-up. From a microbiological perspective, the patient was considered to be cured when M tuberculosis sample culture became sterile. Failure was diagnosed when the culture remained positive after 3 months or more of treatment, and relapse occurred when the culture was again positive after being sterile. Those patients treated for fewer than 6 months were excluded from the efficacy of treatment analysis.
Crude mortality was defined as all deaths occurring during follow-up. Mortality was considered to be related to tuberculosis when death occurred during the course of treatment and there was microbiological or histological evidence of active tuberculosis at the moment of death.
Adverse reactions to antituberculous drugs requiring discontinuation of therapy (at the physician's discretion) were considered to be severe; the remaining adverse reactions were considered to be mild.
Statistical analysis. We compared the baseline characteristics of: (1) the kidney, liver, and heart transplant recipients; (2) patients with pulmonary, extrapulmonary, and disseminated tuberculosis; and (3) patients developing early and late tuberculosis, using Student's t test for quantitative variables, or the chi-square or Fisher's exact test for qualitative variables. The factors influencing crude mortality and mortality directly related with tuberculosis were assessed by means of an univariate analysis. All statistical associations were analyzed using two-tailed t tests. P-values of <0.05 were considered significant.
RESULTS
During the period studied (1980 to 1994), a total of 6326 patients received a solid-organ transplant (kidney, n=4539; liver, n=1202; heart, n=585) in seven reference tertiary centers for transplantation from different geographical areas in Spain. Among these patients, 51 (0.8%) developed tuberculosis (33 renal transplant recipients, 12 liver transplant recipients, and 6 heart transplant recipients). Forty-five of these patients met the criteria for certain tuberculosis (culture of M tuberculosis), and six patients met criteria for possible tuberculosis.
The global incidence of tuberculosis remained stable during the period analyzed (Table 1), and this tendency was uniform along the study in all the hospitals. The incidence of tuberculosis in patients with heart and liver transplants (1%) was slightly higher than that in renal transplant recipients (0.7%), but it did not reach statistical significance (P=0.3).
Clinical features. The main clinical characteristics of our patients are shown in Table 2. There were 30 men and 21 women, with a mean age of 45.3 years (range, 23-67 years). Three patients had received a previous transplant. Infection by HIV was acquired inadvertently from the donor in one renal transplant; there were no other HIV-positive patients in this series. Twelve patients (23%) had clinical and/or radiological evidence of previous tuberculosis (residual tuberculosis in preoperative chest x-ray, n=9; prior history of tuberculosis, n=5). One patient had close contact with a family member with pulmonary tuberculosis. PPD skin testing was performed before transplantation in 30 patients (59%). The result was positive in six patients (20%), but only two of them received prophylaxis against tuberculosis with isonicotinic acid hydrazide (INH). The PPD skin test was negative in the remaining 24 patients, but delayed-type hypersensitivity skin testing was performed in only four of them, and all were anergic. None of the patients who had negative, anergic, or unknown skin tests received antituberculous prophylaxis.
The time that elapsed from the date of transplantation to the onset of the symptoms ranged from 15 days to 13 years. There was a tendency for heart transplant recipients to develop tuberculosis more precociously (mean time, 8.6 months) than liver or kidney transplant recipients (mean time, 18.6 and 27.6 months, respectively; P=0.3). We observed a bimodal distribution in time: 61% of the patients developed tuberculosis early (during the first year after transplantation), all of them in the first 9 months, whereas 39% developed late tuberculosis, all cases occurring 2 or more years after transplantation. We did not find any differences between the early and late groups (data not shown) in relation to age, type of transplanted organ, use of overimmunosuppression (steroids boluses or antilymphocyte antibodies for rejection before diagnosis of tuberculosis), type of tuberculosis, incidence and severity of toxicity due to antituberculous drugs, rejection related to the concomitant use of CsA and rifampin, and crude or related mortality. However, patients who had clinical and/or radiological evidence of prior tuberculosis (n=12) developed tuberculosis after transplantation earlier (265±525 days) than patients (n=39) without antecedents (829±1020 days; P=0.01).
Pulmonary tuberculosis was diagnosed in 32 patients (63%), disseminated disease was diagnosed in 13 patients (25%), and extrapulmonary disease was diagnosed in 6 patients (12%). The percentage of kidney, liver, and heart transplant recipients with pulmonary, disseminated, and extrapulmonary tuberculosis was similar among the three groups. The incidence of disseminated tuberculosis was similar among patients who received overimmunosuppression before the development of tuberculosis and those who did not (27% vs. 22%, P>0.05). Overall, the lung was the single site most frequently involved (42 patients); other organs less frequently involved are shown in Table 2. Chest x-ray showed abnormalities in 36 patients (71%; lung infiltrates, n=18; pleural effusions, n=10; miliary pattern, n=9; solitary pulmonary nodule, n=5; and mediastinal lymphadenopathy, n=3).
Five patients were asymptomatic at the time of diagnosis. In three of them, diagnosis was achieved accidentally as part of a routine surveillance study for tuberculosis. The other two patients were diagnosed on the basis of radiological abnormalities on chest x-ray. In the 46 symptomatic patients, the time from onset of the symptoms to diagnosis ranged from 2 to 180 days (mean time, 41 days). In 17 patients (33%), diagnosis was not suspected initially, and in three patients diagnosis was made at necropsy. In 30 patients (59%), aggressive techniques (fiberoptic bronchoscopy, mediastinoscopy, laparoscopy, tissue biopsies) were needed to reach the diagnosis. Symptoms included fever (n=40), cough (n=18), constitutional syndrome (n=24), osteoarticular pain (n=4), pleuritic chest pain (n=5), shortness of breath (n=6), and acute abdominal pain (n=5).
Treatment and outcome. Of the 51 patients, only 39 received 3 or more months of antituberculous treatment because 5 patients died before diagnosis and 7 died during the first 3 months of treatment. The initial treatment was a three-drug regimen in 28 patients (INH, n=28 patients; ethambutol, n=26; rifampin, n=20; pyrazinamide, n=8; and streptomycin, n=2), and a four-drug regimen in 18 patients (INH, n=18 patients; rifampin, n=18; ethambutol, n=17; pyrazinamide, n=17; and streptomycin, n=2). The initial treatment needed to be modified in seven patients due to drug toxicity, and in two of these patients, more than four drugs were used. In 26 patients, CsA was used simultaneously with rifampin; as a consequence of this association, CsA levels decreased dramatically in all of them (mean decrease of basal CsA levels ±SD, 227±134 ng/ml), and it was necessary to increase greatly the dosage of this drug to maintain its serum levels in a therapeutic range (mean increase of CsA dosage ±SD, 1842±3828 mg/day). Despite these measures, it was necessary to stop the administration of rifampin in seven patients. The 26 patients receiving rifampin and CsA had a higher incidence of acute rejection during treatment (35%) than the remaining group of 20 treated patients who did not receive this combination (19%), although the difference was not significant (P=0.06).
Toxicity secondary to the use of antituberculous drugs was present in 16 of 46 patients (35%) (Table 3). Hepatotoxicity occurred in 15 patients, neurotoxicity occurred in two patients, and cutaneous toxicity occurred in one patient. Two patients suffered from two adverse effects. Hepatotoxicity was severe, as previously defined, in seven patients and mild in eight patients. Six of the 28 patients (21%) receiving three drugs developed hepatotoxicity, in contrast to 9 of 18 patients (50%) who received four drugs (P=0.03). Hepatotoxicity was severe in 1 of 28 patients receiving three drugs (4%) in contrast to 6 of 18 patients (33%) receiving four drugs (P<0.01). One half of the liver transplant recipients receiving antituberculous treatment developed hepatotoxicity that was severe in 25% of cases. The kidney transplant recipients developed hepatotoxicity in 37% of cases, and it was severe in 11% of them. In contrast, none of the heart transplant recipients developed toxicity secondary to antituberculous therapy.
Microbiological, clinical, and graft outcome are also shown in Table 3. Microbiological response was evaluated in the 39 patients who completed 3 or more months of treatment. It was favorable in all except two patients. The first patient was a kidney transplant recipient with pulmonary tuberculosis who developed severe hepatotoxicity and interference with CsA after use of rifampin, which was withdrawn. Treatment was continued with INH, pyrazinamide, and streptomycin, but M tuberculosis continued being cultured from sputum after 6 months of treatment. The patient died shortly after 6 months of treatment, although ethambutol and ofloxacin had been added to the treatment regimen. There was evidence of active tuberculosis in autopsy. Unfortunately, the susceptibility of this strain was not studied. The second patient was a heart transplant recipient who developed pulmonary tuberculosis caused by a strain of M tuberculosis resistant to five conventional antituberculous drugs and died as a consequence of tuberculosis, despite a 6-month period of treatment; susceptibility studies were known after his death. Both patients had antecedents of tuberculosis that had been incorrectly treated, and neither of them had received INH prophylaxis. There were no cases of relapses in the group of 35 patients who completed 6 or more months of treatment.
More than a quarter of patients lost their grafts, generally as a consequence of the interference between CsA and rifampin. The percentage of patients losing their grafts was similar among kidney, liver, and heart transplant recipients. Sixteen patients (31%) died during follow-up. Death was related to tuberculosis in 10 of them (20%), and was unrelated in the remaining six patients. The causes of death related to tuberculosis were: acute respiratory insufficiency during the course of miliary tuberculosis (n=6); severe hepatotoxicity secondary to antituberculous drugs, with evidence of tuberculosis in autopsy (n=2); colon perforation due to intestinal tuberculosis (n=1); and acute rejection due to interference between CsA and rifampin, with evidence of tuberculosis at autopsy (n=1). There was no statistically significant difference in the percentage of crude and tuberculosis-related mortality among kidney, liver, and heart transplant recipients.
Risk factors influencing mortality. In Table 4, several risk factors influencing crude and tuberculosis-related mortality are analyzed. Univariate study showed that patients receiving both boluses of steroids and antilymphocyte antibodies for acute rejection had a crude mortality rate higher than that of patients not receiving antirejection drugs or receiving only boluses of steroids (59% vs. 22%, P<0.05). Rates of crude and tuberculosis-related mortality were greater among those patients receiving treatment for fewer than 9 months than among patients who received treatment for 9 months or more (50% vs. 7% and 33% vs. 0%, respectively, P<0.05). In addition, crude and tuberculosis-related mortality rates were higher in patients who developed acute rejection related to the simultaneous use of CsA and rifampin than in patients who did not develop rejection with this association (56% vs. 18% and 44% vs. 6%, respectively, P=0.03). Finally, patients suffering from other opportunistic infections concomitant with tuberculosis had a higher risk of crude and tuberculosis-related mortality than patients without this complication (58% vs. 24% and 42% vs. 13%, respectively, P<0.05). A multivariate logistic regression analysis was not performed because the number of cases in this series was insufficient.
DISCUSSION
The appearance of the AIDS pandemic has increased the interest in tuberculosis in immunocompromised patients. Curiously, the information available about tuberculosis in other immunosuppressed patients, such as organ transplant recipients, is scarce. Although one can find many case reports of tuberculosis, especially in renal transplant recipients, there are only a few collected series of tuberculosis in organ transplant recipients (3-13). In Table 5, the more important data on 272 transplant patients with tuberculosis (including ours) published in the last 30 years in series presenting at least four cases are summarized. Our study was carried out with data obtained from 51 kidney, liver, and heart transplant recipients collected over 15 years, and it is the largest series published so far in the literature.
The incidence of tuberculosis among kidney, liver, and heart transplant recipients was 0.8%, a figure 20-fold higher than our national average for the general population, which is very high in Spain and has been estimated to be 30-50 cases per 100,000 (14, 15). It is noteworthy that although the incidence of tuberculosis increased in our country during the study period (16, 17), the incidence of tuberculosis in organ transplant recipients has remained unmodified during this time. This could be explained by the fact that the higher incidence of tuberculosis in our country has been substantially due to the increasing number of AIDS cases; in one of our patients, tuberculosis related to AIDS. The incidence reported here is similar to that found in another series from Spain dealing with tuberculosis in transplantation (18, 19).
The reported frequency of infection by M tuberculosis in the transplant population (Table 5) ranges from 0.2% to 15% (mean frequency, 3.7%), which is 6-62 (mean, 26) times the incidence in the general population (20-24). Geographic differences may influence this wide range of figures. In North America, for example, the frequency ranges from 0.5% to 1% (3, 25), in Northern Europe it is 1-4% (7, 20, 26), and in areas such as India and Pakistan it is over 9% (27, 28). There does not appear to be any difference in the incidence of tuberculosis found among the three types of solid organ transplants studied, although kidney transplants had the lowest incidence (Table 1). The incidence of tuberculosis in recipients of other transplanted organs (pancreas, bone marrow, heart-lung, and lung) must be low, but it is less well known and only a few cases have been published so far (1, 29).
The time of onset of symptoms after transplantation varies; however, we and other authors have observed a bimodal distribution (4, 10). The majority (61%) of our cases appeared early (during the first year after transplantation), and the remaining 39% of cases appeared late (after the first year of transplantation). These data agree with reports indicating that the majority of Mycobacterium infections occur within the first 12 months after transplantation. Although it has been suggested that patients who develop early tuberculosis are more severely immunosuppressed than patients with late tuberculosis (30), we and other authors (10) did not find any differences between these two groups (data not shown). In contrast, patients with prior clinical or radiological evidence of tuberculosis developed the disease earlier than did patients without these antecedents. These data suggest that patients with a history of tuberculosis have a higher risk of reactivation during the first months after transplantation, independent of the type of immunosuppression received. A recent study indicates, however, that patients receiving CsA developed tuberculosis earlier (mean time, 5.5 months) than did patients on conventional immunosuppression (mean time, 24 months; P<0.01) (11).
This series and others show that tuberculosis following solid-organ transplantation is most frequently limited to the lung and less frequently disseminated and extrapulmonary (Table 5). However, disseminated and extrapulmonary disease occurs more commonly than in the general population (25), with incidence rates as high as 38-64% (4, 25). Of note was the high number of patients with miliary tuberculosis who developed initial symptoms resembling life-threatening sepsis or adult respiratory distress syndrome. Patients with early tuberculosis or those receiving overimmunosuppression (boluses of steroids or antilymphocyte antibodies) do not appear to be at higher risk for the development of disseminated disease.
Fever, leukocytosis, night sweats, weight loss, and lymphadenopathy were the most frequent initial manifestations of tuberculosis in this and other series (31). Importantly, the disease was asymptomatic in 10% of cases. Although sputum smears for acid-fast bacilli were positive in 67% of patients and 71% had abnormalities on chest x-ray, suspicion of tuberculosis was not aroused, and it took a mean time of 41 days for diagnosis. It was necessary to perform an aggressive diagnostic maneuver in 59% of cases; a similar percentage has been communicated in other series (4). This indicates that physicians do not usually consider the diagnosis of tuberculosis in transplant patients; a major index of suspicion appears to be necessary in these cases. Invasive diagnostic methods, such as bronchoscopy or lung biopsy, should be performed early in solid-organ transplant recipients with suspected tuberculosis. In three of our patients, diagnosis was achieved by means of a routine sputum culture as part of a surveillance protocol for early diagnosis of tuberculosis, carried out in one of the participating hospitals. This approach could be suitable for rapid diagnosis of tuberculosis in transplant patients (32). Polymerase chain reaction techniques may play a role in future screening protocols for donors and recipients.
Tuberculosis has important implications in the outcome of transplant patients. Thirty-one percent of our patients died during follow-up. In 20%, death was directly related to tuberculosis (there was evidence of active tuberculosis at time of death), although in several of these patients there were other causes that could have contributed to death. The average percentage of mortality directly related to tuberculosis in several large series of transplant recipients reviewed was 15% (Table 5). Tuberculosis causes tremendous mortality in these patients as compared with mortality in the general population and in transplant patients without tuberculosis (3, 23, 33). It is of concern that this high mortality rate has not varied in the more recently published series, despite the decreasing incidence of tuberculosis recorded in many countries (Table 5).
The important drug interactions associated with antimy-cobacterial therapy in solid-organ transplant recipients are relatively unique and play a considerable role in the poor outcome of tuberculosis in this population. Rifampin increases the catabolism of corticosteroids and CsA, making it impossible to maintain therapeutic serum CsA concentrations, despite massive dose increases (34, 35). The simultaneous use of CsA and rifampin dramatically lowered CsA levels in our patients, and favored the development of acute rejection in 35% of patients receiving this combination. Importantly, 27% of our patients lost their grafts due to rejection, and in the majority of cases, rejection was due to the interference of rifampin with CsA. Other authors have reported similar findings (36-40). Rejection following interference of CsA and rifampin was one of the most significant risk factors for both crude and tuberculosis-related mortality (Table 4). In our opinion, rifampin use must be avoided in transplant recipients with tuberculosis, even if CsA levels are closely monitored. If rifampin is to be administered, the CsA dose should initially be increased three- to fivefold, glucocorticoid doses should be doubled, the frequency of administration of CsA should be increased from twice to three times daily, and serum levels should be monitored daily until they stabilize (41). Tacrolimus (FK506) could be an alternative to CsA in these cases. Alternatively, CsA use may be avoided, but this is a less advisable measure. Other risk factors associated with higher mortality were the use of antilymphocyte antibodies and the development of opportunistic infections together with tuberculosis. Disseminated tuberculosis does not appear to have an unfavorable prognosis.
Ninety-five percent of patients treated for 6 months or more achieved microbiological cure; therefore, treatment appears to be equally effective in transplant patients and in the general population. Our data and the data published by others authors suggest that it is not necessary to reduce immunosuppression to improve prognosis if appropriate therapy is instituted promptly (42, 43). There is no consensus with regard to optimal treatment in these patients. Several authors have suggested that a course of 6-9 months is effective, at least in patients with more localized disease (7, 10, 44); however, a longer course has been recommended by others (5, 22, 45). In our experience, mortality was lower in patients receiving more than 9 months of treatment, although a 6-month regimen could be equally effective in some cases. In our two patients who died after 6 months of treatment, there was microbiological evidence or clinical suspicion that disease was due to multiresistant isolates of M tuberculosis, so that a 9-month course would be equally ineffective. In general, more prolonged treatments seem to be necessary, as rifampin cannot be used in the majority of cases, and alternative drugs with a lower activity against M tuberculosis must be given.
It appears reasonable that toxicity due to antituberculous drugs should be greater in transplant recipients, as has been described (8, 46). One third of our patients developed some degree of drug toxicity during treatment, a figure higher than in the general population (44). Hepatotoxicity was the most frequent adverse effect, and it was observed only in liver and kidney recipients. Hepatotoxicity is especially disturbing in liver transplantation because it may be difficult to differentiate from rejection. Hepatotoxicity was generally severe in our series, requiring definitive cessation of treatment in 14% of patients; a similar incidence has been reported by other authors (Table 5); this figure may be an underestimation, since many series provide scant information on the incidence and severity of hepatotoxicity (Table 5).
Prevention of tuberculosis in organ transplantation is a complex issue. Patients are often anergic before transplantation due to the underlying disease, thereby making preoperative screening with PPD skin tests frequently unrevealing. This may explain why pretransplantation PPD and delayed-type hypersensitivity skin tests were not generally used in the present series. Indeed, PPD skin testing was performed in only 59% of our patients, and delayed-type hypersensitivity skin testing was used in only 8%. Prophylaxis with INH in this population is a controversial issue due to the high risk of hepatotoxicity. The general tendency in the literature is to avoid the use of INH, even in countries with a high incidence of tuberculosis (4, 10, 12). This tendency is reflected by the low percentage of patients in our series who received prophylaxis, despite its being indicated in nearly one third of our cases. INH prophylaxis would be generally advisable in PPD-positive transplant recipients, since the incidence of tuberculosis in this population is three times the expected incidence in patients with negative skin tests (4, 5). The American Thoracic Society recommends routine administration of INH to tuberculin-positive transplant recipients with healed pulmonary tuberculosis (47), but some authors think that regular use of INH, especially in liver transplantation, is not justified due to the elevated risk of hepatotoxicity (12, 48). In a series of renal patients receiving INH prophylaxis, 11% had evidence of hepatic dysfunction attributable solely to INH, and 2.5% developed major hepatic dys-function, with two deaths related to liver failure (49). Results like these prompt many authors to recommend that INH chemoprophylaxis be limited to patients with a positive tuberculin skin test who fall into high-risk categories (patients born in areas where tuberculosis is endemic, and patients with chronic advanced underlying illness, rejection, and/or adjunctive immunosuppression), have abnormal chest x-rays, or have documented recent skin conversion (50, 51). There are no data addressing prophylaxis if the recipient is tuberculin negative and receives an organ from a PPD-positive donor, but some authors suggest that a course of INH should be considered (4, 23, 52). It would be worthwhile to know whether anergic transplant recipients, such as PPD-positive patients, are at high risk for development of tuberculosis, as are patients with AIDS (53), and whether preventive therapy should be offered to these patients, at least those living in areas with a high incidence of tuberculosis (54). In fact, one analysis to determine the need for prophylaxis in liver transplant patients found the risk factors for developing tuberculosis to be similar to those of HIV infection (55).
Our study had some limitations, the most important being that it was not a comparative case control study, which would have better informed us as to the risk factors for developing tuberculosis and related mortality in solid-organ transplant recipients. The risk factors for tuberculosis in transplant patients are as yet poorly defined, although the role of intensification of immunosuppression for a failing graft appears to be important (10). In fact, 65% of our patients were overimmunosuppressed, a figure clearly higher than that recognized for the general transplant population. Although some authors have suggested that the use of antilymphocyte antibodies (especially OKT3) increases the risk of tuberculosis dissemination (8, 50), as occurs in mice (51), our data do not concur with this suggestion, although crude mortality increased when these drugs were used. This study was not designed to define the optimum length and composition of therapy and prophylaxis for tuberculosis in transplant patients. Although a 6-month course of treatment could be effective, the exclusion of rifampin increases the risk of reactivation and spread of tuberculosis, and a more prolonged course may be advisable. Our results suggest that treatment with three drugs for at least 9 months, avoiding rifampin, reduces the mortality associated with this disease.
In summary, M tuberculosis is clearly a potential pathogen in solid-organ transplant recipients. Tuberculosis is bimodally distributed in the posttransplantation period and usually tends to present as pulmonary or, less frequently, disseminated infection. The resurgence of tuberculosis during the past decade establishes the need to maintain a high index of suspicion, to improve methods for screening, and to initiate prophylaxis in patients at high risk for reactivation of latent infection.
Acknowledgments. The authors thank Dr. Blanca Miranda, Spanish National Organization for Transplantation, Madrid, for providing patient demographic data.
Footnotes
Presented in part at the 34th Interscience Conference on Anti-microbial Agents and Chemotherapy, Orlando, Florida, October 4-7, 1994 (Abstract J116).
Abbreviations: CsA, cyclosporine; INH, isonicotinic acid hydrazide; PPD, purified protein derivative of tuberculin.
REFERENCES
1. Miller RA, Lanza, LA, Kline JN, Geist LJ.
Mycobacterium tuberculosis in lung transplant recipients. Am J Respir Crit Care Med 1995; 152: 374.
2. Roberts G, Koneman E, Yoo KM.
Mycobacterium. In: Balows A, Hausler WJ, Herrmann KL, Isenberg HD, Shadomy HJ, eds. Manual of clinical microbiology, 5th ed. Washington, DC: American Society for Microbiology, 1991: 304.
3. Lloveras J, Phillip K, Peterson R, Simmons RL, Najarian JS. Mycobacterial infections in renal transplant recipients. Arch Intern Med 1982; 142: 888.
4. Qunibi WY, Al-Sibai MB, Taher S, et al. Mycobacterial infection after renal transplantation: report of 14 cases and review of the literature. Q J Med 1990; 77: 1039.
5. Higgins RM, Cahn AP, Porter D, et al. Mycobacterial infections after renal transplantation. Q J Med 1991; 78: 145.
6. Koselj M, Buturovic J, Malovrth M. Tuberculosis in renal allograft recipients. Transplant Proc 1992; 24: 1909.
7. Riska H, Grönhagen-Riska C, Ahonen J. Tuberculosis and renal allograft transplantation. Transplant Proc 1987; 19: 4096.
8. Higgins RSD, Kusne S, Reyes J, et al.
Mycobacterium tuberculosis after liver transplantation: management and guidelines for prevention. Clin Transplant 1992; 6: 81.
9. Delaney V, Sumrani N, Hong JH, Sommer B. Mycobacterial infections in renal allograft recipients. Transplant Proc 1993; 25: 2288.
10. Hall CM, Willcox PA, Swanepoel CR, Kahn D, Van Zyl-Smit R. Mycobacterial infection in renal transplant recipients. Chest 1994; 106: 435.
11. John GT, Vincent L, Jeyaseelan L, Jacob CK, Shastry JCM. Cyclosporine immunosuppression and mycobacterial infections. Transplantation 1994; 58: 247.
12. Stöblen F, Neuhaus R, Neumann K, et al. Tuberculosis in liver transplant recipients: recurrence after transplantation? Transplant Proc 1994; 26: 3604.
13. Wong LL, Levin BS, Collins GM, Bry WI. Unusual manifestations of tuberculosis in cyclosporine-treated renal transplant recipients. Clin Transplant 1993; 7: 18.
14. De March P. Situación actual de la tuberculosis en España. Med Clín 1991; 97: 463.
15. Vidal R, Garcia J, Ruiz J, Fite E, Monsó E, Martín N. Estudio controlado de 637 pacientes con tuberculosis: diagnóstico y resultados terapéuticos con esquemas de 9 y 6 meses. Med Clín 1986; 87: 368.
16. De March P. Tuberculosis y SIDA: situación en España y perspectivas. Rev Clín Esp 1990; 186: 365.
17. Rey R, Ausina V, Casal M, et al. Situación actual de la tuberculosis en España: una perspectiva sanitaria en precario respecto a los países desarrollados. Med Clín 1995; 105: 703.
18. Garcia-Leoni ME, Martin-Scapa C, Rodeño P, Valderrábano F, Moreno S, Bouza E. High incidence of tuberculosis in renal patients. Eur J Clin Microbiol Infect Dis 1990; 9: 283.
19. Muñoz P, Palomo J, Muñoz R, Rodríguez-Creixéms M, Peláez T, Bouza E. Tuberculosis in heart transplant recipients. Clin Infect Dis 1995; 21: 398.
20. Coutts IJ, Jegarajah S, Stark JE. Tuberculosis in renal transplant recipients. Br J Dis Chest 1979; 73: 141.
21. Riska H, Kuhlback B. Tuberculosis and kidney transplantation. Acta Med Scand 1979; 205: 637.
22. Neff TA, Hudgel DW. Miliary tuberculosis in a renal transplant recipient. Am Rev Respir Dis 1973; 108: 677.
23. Sinnott JT, Emmanuel PJ. Mycobacterial infections in the transplant patient. Semin Respir Infect 1990; 5: 65.
24. Delaney V, Sumrani N, Hong JH, Sommer B. Mycobacterial infections in renal allograft recipients. Transplant Proc 1993; 25: 2288.
25. Lichtenstein IH, MacGregor RR. Mycobacterial infections in renal transplant recipients: report of five cases and review of the literature. Rev Infect Dis 1983; 5: 216.
26. McWhinney N, Khan O, Williams G. Tuberculosis in patients undergoing maintenance haemodialysis and renal transplantation. Br J Surg 1981; 68: 408.
27. Malhotra KK, Dash SC, Dhawan IK, Bhuyan UN, Gupta A. Tuberculosis and renal transplantation: observations from an endemic area of tuberculosis. Postgrad Med J 1986; 62: 359.
28. Naqvi SA, Hussain M, Askari H, et al. Is there a place for prophylaxis against tuberculosis following renal transplantation? Transplant Proc 1992; 24: 1912.
29. Dromer C, Nashef SAM, Velly JF, Martigne C, Couraud L. Tuberculosis in transplanted lungs. J Heart Lung Transplant 1993; 12: 924.
30. Jereb JA, Burwen DR, Dooley SW, Haas WH, Crawford JT, Geiter LJ. Nosocomial outbreak of tuberculosis in a renal transplant unit: application of a new technique for restriction fragment polymorphism analysis of
Mycobacterium tuberculosis isolates. J Infect Dis 1993; 168: 1219.
31. Costa JMN, Meyers AM, Botha JR, Conlan AA, Myburgh A. Mycobacterial infections in recipients of kidney allografts: a seventeen-year experience. Acta Med Port 1988; 1: 51.
32. Torre-Cisneros J, De la Mata M, Rufian S, et al. Importance of surveillance mycobacterial cultures after liver transplantation. Transplantation 1995; 60: 1054.
33. Sundberg R, Shapiro R, Darras F, et al. A tuberculosis outbreak in a renal transplant program. Transplant Proc 1991; 23: 3091.
34. Langhoff E, Madsen S. Rapid metabolism of cyclosporin and prednisone in kidney transplant patient receiving tuberculostatic treatment [Letter]. Lancet 1983; 2: 1031.
35. Cassidy MJD, Van Zyl-Smit R, Pascoe MD, Swanepoel CR, Jacobson JE. Effect of rifampicin on cyclosporin A blood levels in a renal transplant recipient [Letter]. Nephron 1985; 41: 207.
36. Offermann G, Keller F, Molzahn M. Low cyclosporine A bloods level and acute graft rejection in a renal transplant recipient during rifampin treatment. Am J Nephrol 1985; 5: 385.
37. Buffington GA, Dominguez JH, Piering WF, Hebert LA, Kauffman HM Jr, Lemann J Jr. Interaction of rifampicin and glucocorticoids: adverse effect on renal allograft function. JAMA 1976; 236: 1958.
38. Modry DL, Stinson EB, Oyer PE, Jamieson SW, Baldwin JC, Shumway NE. Acute rejection and massive cyclosporine requirements in heart transplant recipients treated with rifampin. Transplantation 1985; 39: 313.
39. Chan GL, Sinnott JT, Emmanuel PJ, Cyanle S, Weinstein SS. Drug interactions with cyclosporine: focus on antimicrobial agents. Clin Transplant 1992; 6: 141.
40. Daniels NJ, Dover JS, Schachter RK. Interaction between cyclosporin and rifampin [Letter]. Lancet 1984; 2: 639.
41. Al-Sulaiman MH, Dhar JM, Al-Khader AA. Successful use of rifampin in the treatment of tuberculosis in renal transplant patients immunosuppressed with cyclosporine. Transplantation 1990; 50: 597.
42. Ettinger NA, Trulock EP. Pulmonary considerations of organ transplantation. Am Rev Respir Dis 1991; 143: 1386.
43. Abbott MR, Smith DD. Mycobacterial infections in immunosuppressed patient. South Med J Aust 1981; 1: 351.
44. American Thoracic Society. Medical Section of the American Lung Association: treatment of tuberculosis and tuberculosis infection in adults and children. Am Rev Respir Dis 1986; 134: 355.
45. Al Batata M. Pulmonary tuberculosis in a renal transplant recipient. JAMA 1977; 237: 1465.
46. Thomas PA Jr, Mozes MF, Jonasson O. Hepatic dysfunction during isoniazid chemoprophylaxis in renal allograft recipients. Arch Surg 1979; 114: 597.
47. American Thoracic Society. Preventive therapy of tuberculous infection. Am Rev Respir Dis 1974; 110: 371.
48. John GT, Thomas PP, Thomas M, Jeyaseelan L, Jacob CK, Shastry JCM. A double-blind randomized controlled trial of primary isoniazid prophylaxis in dialysis and transplant patients. Transplantation 1994; 57: 1683.
49. Thomas PA, Manko MA. Chemoprophylaxis for the prevention of the tuberculosis in the immunosuppressed renal allograft recipient. Transplantation 1975; 20: 76.
50. Meyers BR, Halpern M, Sheiner P, Mendelson MH, Neibart E, Miller C. Tuberculosis in liver transplant patients. Transplantation 1994; 58: 301.
51. Mullerova M, Pekarek J, Nouzak J. Immunosuppression and experimental tuberculosis. Biomedicine 1974; 20: 390.
52. Gottesdiener KM. Transplanted infections: donor-to-host transmission with the allograft. Ann Intern Med 1989; 110: 1001.
53. Moreno S, Baraia-Etxaburu J, Bouza E, et al. Risk for developing tuberculosis among anergic patients infected with HIV. Ann Intern Med 1993; 119: 194.
54. Kusne S, Irish W, Geary K, Ondick L, Rakela J, Fung J. Anergy panel and purified protein derivative skin testing are poor tuberculosis screening methods in adult liver candidates [Abstract J33]. In: Program and abstracts: 35th Interscience Conference on Antimicrobial Agents and Chemotherapy. San Francisco: American Society for Microbiology, 1995: 262.
55. Passannante MR, Restifo RA, Reichman LB. Preventive therapy for the patient with both universal indication and contraindication for isoniazid. Chest 1993; 103: 825.
56. Sakhuja V, Jha V, Varma PP, Joshi K, Chugh KS. The high incidence of tuberculosis among renal transplant recipients in India. Transplantation 1996; 61: 211.
