Chest computed tomography is more likely to show latent tuberculosis foci than simple chest radiography in liver transplant candidates

Authors

  • Jiwon Lyu,

    1. Division of Pulmonary and Critical Care Medicine, University of Ulsan College of Medicine, Asan Medical Center, Songpa-Gu, Seoul, South Korea
    Search for more papers by this author
  • Sung-Gyu Lee,

    1. Division of Hepatopancreatobiliary Surgery and Liver Transplantation, Department of Surgery, University of Ulsan College of Medicine, Asan Medical Center, Songpa-Gu, Seoul, South Korea
    Search for more papers by this author
  • Shin Hwang,

    1. Division of Hepatopancreatobiliary Surgery and Liver Transplantation, Department of Surgery, University of Ulsan College of Medicine, Asan Medical Center, Songpa-Gu, Seoul, South Korea
    Search for more papers by this author
  • Sang-Oh Lee,

    1. Department of Infectious Diseases, University of Ulsan College of Medicine, Asan Medical Center, Songpa-Gu, Seoul, South Korea
    Search for more papers by this author
  • Oh-Hyun Cho,

    1. Department of Infectious Diseases, University of Ulsan College of Medicine, Asan Medical Center, Songpa-Gu, Seoul, South Korea
    Search for more papers by this author
  • Eun Jin Chae,

    1. Department of Radiology and Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, Songpa-Gu, Seoul, South Korea
    Search for more papers by this author
  • Sang Do Lee,

    1. Division of Pulmonary and Critical Care Medicine, University of Ulsan College of Medicine, Asan Medical Center, Songpa-Gu, Seoul, South Korea
    Search for more papers by this author
  • Woo Sung Kim,

    1. Division of Pulmonary and Critical Care Medicine, University of Ulsan College of Medicine, Asan Medical Center, Songpa-Gu, Seoul, South Korea
    Search for more papers by this author
  • Dong Soon Kim,

    1. Division of Pulmonary and Critical Care Medicine, University of Ulsan College of Medicine, Asan Medical Center, Songpa-Gu, Seoul, South Korea
    Search for more papers by this author
  • Tae Sun Shim

    Corresponding author
    1. Division of Pulmonary and Critical Care Medicine, University of Ulsan College of Medicine, Asan Medical Center, Songpa-Gu, Seoul, South Korea
    • Division of Pulmonary and Critical Care Medicine, University of Ulsan College of Medicine, Asan Medical Center, 388-1 Poongnap-Dong, Songpa-Gu, Seoul, South Korea 138-736
    Search for more papers by this author
    • Telephone: 82-2-3010-3892; FAX: 82-2-3010-6968


Abstract

Although the detection and treatment of latent tuberculosis infections (LTBIs) in transplant candidates are essential, current diagnostic methods for LTBIs are limited, especially in immunocompromised subjects. Pretransplant chest computed tomography (CT) may reveal more LTBI foci and thus predict the development of posttransplant tuberculosis (TB) more efficiently; however, this hypothesis has not yet been investigated. Thirty-six liver transplantation (LT) recipients who developed TB (the TB group) and 144 LT recipients who did not develop TB (the control group) were retrospectively enrolled into a study with a nested case-control design, and their clinical characteristics and radiological findings were compared. Tuberculin skin tests (TSTs) were not performed, and none of these patients had been treated for LTBIs. Thirty-six of 2549 LT recipients (1.4%) were diagnosed with TB after LT (median = 10 months, range = 1-80 months). Twenty-eight patients (77.8%) successfully completed the treatment. There were no significant differences in the clinical characteristics of the 2 groups. Abnormal CT findings (40.0% versus 17.3%, P = 0.018) and chest X-ray (CXR) findings (25.0% versus 11.8%, P = 0.044) suggestive of healed TB were significantly more frequent in the TB group versus the control group. Of the 10 patients who underwent chest CT and developed TB, 5 (50%) showed abnormal findings only on chest CT scans, whereas their CXR results were normal. In conclusion, a pretransplant chest CT scan is more likely to show an LTBI than a CXR in those with post-LT TB. The usefulness of chest CT along with traditional methods such as TSTs for LTBI screening should be further investigated. Liver Transpl 17:963–968, 2011. © 2011 AASLD.

The incidence of tuberculosis (TB) has been estimated to be 20 to 74 times higher in transplant recipients versus the general population; moreover, TB in transplant recipients is associated with mortality rates as high as 30%.1, 2 TB in transplant recipients is usually caused by the reactivation of a latent tuberculosis infection (LTBI) rather than the acquisition of a new TB infection.1, 3 Therefore, the diagnosis and treatment of LTBIs are generally recommended for transplant candidates.4, 5 The prevalence of TB infections is especially high in liver transplantation (LT) recipients; rates of 1% to 6% have been reported in some case series, and they may be even higher in areas endemic for TB.1, 2, 6, 7 Furthermore, the mortality rates of LT recipients with LTBIs have been reported to be as high as 30% to 100%.1, 8, 9

Despite the high risk of active TB in LT recipients, there is no consensus regarding the appropriate diagnostic and treatment methods for LTBIs.10 Although tuberculin skin tests (TSTs) are generally used to screen for TB infections, they do not always diagnose LTBIs in LT patients. Moreover, according to a previous report,11 the highest risk factor for active posttransplant TB is radiological evidence of a previous TB infection, not positive TST results. Because LT recipients with radiological features of healed TB are at high risk for active posttransplant TB, they should be treated for LTBIs regardless of their TST results.1 No studies to date have assessed the usefulness of chest computed tomography (CT) in screening LT candidates for LTBIs. Therefore, we retrospectively analyzed the ability of pre-LT chest CT scans to predict the development of post-LT TB.

Abbreviations:

CNI, calcineurin inhibitor; CT, computed tomography; CXR, chest X-ray; IGRA, interferon-γ release assay; LT, liver transplantation; LTBI, latent tuberculosis infection; TB, tuberculosis; TST, tuberculin skin test.

PATIENTS AND METHODS

Patients

Patients who underwent primary LT between January 1999 and December 2009 at the Asan Medical Center and survived more than 1 month were retrospectively evaluated until August 2010. One hundred five of the 2654 LT recipients died within 30 days; thus, 2549 patients were eligible for this study. Thirty-six of these patients developed TB. Using a nested case-control design, we selected 4 control subjects who did not develop TB and matched those patients who developed TB with respect to sex and age. The compared factors included demographic features, comorbidities, a previous history of TB, simple chest X-ray (CXR) and CT findings, the date of LT, the occurrence of allograft rejection, coinfections with cytomegalovirus, a TB diagnosis, the time from LT to the TB diagnosis, the date of the last follow-up visit, and the date of death. This study was approved by the institutional review board of the Asan Medical Center.

Immunosuppressive Therapy Protocol

The peritransplant primary immunosuppression protocols used for adult LT recipients at our institution consisted of an interleukin-2 receptor inhibitor (basiliximab) on days 0 and 4, an intraoperative steroid bolus (5-10 mg/kg), an intravenous or oral calcineurin inhibitor (CNI), and corticosteroid recycling (which was begun on day 1); adjunctive mycophenolate mofetil was given to patients who showed CNI-associated side effects and to patients whose immunosuppressive treatment needed to be augmented. The oral CNI was usually tacrolimus. Patients were tapered off the steroid rapidly. Patients showing an intractable intolerance of CNIs were given mycophenolate mofetil monotherapy.

Definition of the TB Diagnosis

TB was diagnosed when (1) Mycobacterium tuberculosis or acid-fast bacilli were detected in any clinical specimen, (2) any clinical specimen showed a positive polymerase chain reaction result for M. tuberculosis, (3) a caseating granuloma was found in the tissue, or (4) the clinical findings were compatible with TB and improved after an anti-TB treatment.

Chest Radiography and CT Scanning

Pre-LT standard posteroanterior and lateral CXRs and CT results, if they were available, were reanalyzed for abnormal lesions suggestive of healed TB and for normal or nonspecific abnormal findings. All CXRs and CT results were analyzed by a radiologist (E.J.C.) and a pulmonologist (J.L.) blinded to each patient's clinical history. CXRs were evaluated for fibrotic linear opacities and calcified nodules, and CT findings were evaluated for fibrotic irregular lines, uncalcified nodules, calcified nodules, and fibrotic consolidation. For patients with TB, chest CT findings at the time of diagnosis were compared with baseline CT findings. After they reviewed the CT findings, the 2 observers reached a final diagnosis by consensus.

Assessment of the Treatment Outcomes

The treatment outcomes (cure, treatment completion, failure, default, and death) were based on the guidelines of the World Health Organization.12 Treatment success was defined as the achievement of both a cure and treatment completion.

Statistical Analysis

Results are presented as means and standard deviations. Differences between categorical variables were assessed with the chi-square test or Fisher's exact test (as appropriate), and differences between continuous variables were evaluated with unpaired t tests. A P value <0.05 was considered statistically significant. All statistical analyses were performed with SPSS 13 for Windows (SPSS, Inc., Chicago, IL).

RESULTS

Baseline Clinical Characteristics

Thirty-six of the 2549 LT recipients (1.4%) developed active TB within a median follow-up period of 45 months (range = 4-130 months). Table 1 shows the baseline clinical characteristics of the study participants. TSTs and LTBI prophylaxis were not routinely used. All participants were negative for human immunodeficiency virus. The mean age of the patients with TB was 51.1 ± 9.9 years. Six patients in the TB group (16.7%) and 13 in the control group (9.0%) had a previous history of treatment for active TB disease (P = 0.182). All other baseline characteristics were similar in the 2 groups. The medical records did not describe the previous history of anti-TB treatment in detail, so we did not have enough information about the number of anti-TB agents, the duration of therapy, and noncompliance issues. Instead, active TB was thoroughly excluded before transplantation.

Table 1. Baseline Characteristics of the LT Recipients
CharacteristicTB Group (n = 36)Control Group (n = 144)P Value
  • NOTE: The data are presented as numbers and percentages unless otherwise noted.

  • *

    The data are presented as means and standard deviations.

Age (years)*51.1 ± 9.950.6 ± 9.60.981
Male26 (72.2)104 (72.2)1.000
Indication for LT   
 Hepatocellular carcinoma19 (52.8)57 (39.6)0.152
 Hepatitis B15 (41.7)65 (45.1)0.708
 Other2 (5.6)22 (15.3)0.172
Potential risk factors for TB   
 Diabetes7 (19.4)24 (16.7)0.693
 Previous TB history6 (16.7)13 (9.0)0.182
 CXR suggestive of healed TB9 (25.0)17 (11.8)0.044
 CT suggestive of healed TB10/25 (40.0)14/81 (17.3)0.018
 Allograft rejection6 (16.7)23 (16.0)0.919
 Cytomegalovirus coinfection1 (2.8)2 (1.4)0.490
 Induction immunosuppression22 (61.1)109 (75.7)0.079

Clinical Course and Development of TB

The median time from LT to the TB diagnosis was 10 months (range = 1-80 months). Twenty-one patients (58.3%) were diagnosed at 0 to 6 months, 10 (27.8%) were diagnosed at 7 months to 1 year, and 5 (13.9%) were diagnosed after 1 year. Twenty-three of these 36 patients (63.9%) were diagnosed with pulmonary TB, and 13 (36.1%) were diagnosed with extrapulmonary TB.

All LT recipients received an anti-TB treatment for a mean duration of 10 months with less hepatotoxic drugs; 34 patients (94.4%) were treated with ethambutol, and 27 (75.0%) were treated with fluoroquinolone. Thirty patients (83.3%) received isoniazid, and 14 (38.9%) received rifampin. Rifabutin was used in 11 patients (30.6%) in order to reduce the drug interactions with CNIs. Patients showing evidence of hepatotoxicity were switched to nonhepatotoxic agents. The treatment success rate was 77.8%. Two patients (5.6%) died, with 1 death (2.8%) related to active TB. The clinical characteristics and outcomes of the LT recipients with TB are shown in Table 2.

Table 2. Characteristics of the 36 Patients Who Developed TB After LT
CharacteristicPatients [n (%)]
  • *

    Three of the five patients had TB pleurisy with lymphocyte-dominant exudates and high adenosine deaminase concentrations.

  • One death was related to TB, and 1 death was related to hepatocellular carcinoma.

Location of TB 
 Pulmonary23 (63.9)
 Extrapulmonary13 (36.1)
  Miliary8 (22.2)
  TB pleurisy3 (8.3)
  TB lymphadenopathy1 (2.8)
  TB meningitis1 (2.8)
Diagnostic methods for TB 
 Acid-fast bacillus smear or culture (positive)27 (75.0)
 Polymerase chain reaction for M. tuberculosis (positive)1 (2.8)
 Histology3 (8.3)
 Clinical*5 (13.9)
Drug susceptibility results22/27 (81.5)
 Pansusceptible19/22 (86.4)
 Resistance to isoniazid or streptomycin3/22 (13.6)
TB treatment regimen 
 Intensive phase 
  3 drugs12 (33.3)
  4 or more drugs24 (66.7)
 Continuation phase 
  2 drugs3 (8.3)
  3 drugs14 (38.9)
  4 drugs19 (52.8)
Treatment outcomes 
 Treatment success28 (77.8)
 Death2 (5.6)
 Current treatment3 (8.3)
 Transfer3 (8.3)

Chest Radiography and CT Scans

All subjects in both groups underwent chest radiography, whereas only 25 TB subjects (69.4%) and 81 control subjects (56.3%) underwent chest CT scans. The most common reason for the latter was a metastatic workup for underlying hepatocellular carcinoma.

Abnormal CXRs suggestive of healed TB were observed in 9 TB subjects (25.0%) and 17 control subjects (11.8%, P = 0.044). In addition, CT scans showed that abnormal lesions suggestive of healed TB were more frequent in the TB group versus the control group (40.0% versus 17.3%, P = 0.018; Table 1). CXR and chest CT findings did not differ significantly between the 2 groups except for fibrotic consolidation on chest CT scans (P = 0.011; Table 3). Seven of the 10 patients (70%) who underwent chest CT and developed TB after LT showed TB reactivation at the site of the healed TB. Moreover, 5 of these 10 patients (50%) showed abnormal findings suggestive of healed TB on chest CT scans, but they had normal chest radiography results (Table 4 and Fig. 1).

Table 3. Pretransplant CXR and CT Findings in LT Recipients
CharacteristicTB Group (n = 36)Control Group (n = 144)P Value
  1. NOTE: The data are presented as numbers and percentages.

CXR suggestive of healed TB9 (25.0)17 (11.8)0.044
 Fibrotic linear opacities7 (19.4)15 (10.4)0.157
 Calcified nodules7 (19.4)12 (8.3)0.068
CT suggestive of healed TB10/25 (40.0)14/81 (17.3)0.018
 Irregular lines6 (24.0)11 (13.6)0.224
 Uncalcified nodules3 (12.0)4 (4.9)0.352
 Calcified nodules8 (32.0)11 (13.6)0.069
 Fibrotic consolidation4 (16.0)1 (1.2)0.011
Table 4. Sites of TB in 10 Patients With Healed TB Lesions on Chest CT Images
PatientPre-LT Presence of Healed TBSites of TB on Chest CT Images
CXRChest CTPre-LTPost-LTSame Site
1YesYesRight upper lobeRight upper lobeYes
2NoYesLeft upper lobeLeft upper lobeYes
3YesYesBoth upper lobesRight upper lobeYes
4NoYesRight upper lobeRight upper lobeYes
5YesYesLeft upper lobeMiliary TBNo
6YesYesRight upper lobeRight upper lobeYes
7NoYesRight upper lobeRight upper lobeYes
8YesYesBoth upper lobesLeft upper lobeYes
9NoYesRight lower lobeRight upper lobeNo
10NoYesRight upper lobeMiliary TBNo
Figure 1.

Representative chest radiography and CT images. (A) A pretransplant CXR appeared to be normal, but (B) pretransplant chest CT scanning revealed a TB-suggestive lesion (an uncalcified nodule). (C) Active TB developed 6 months after LT in the same location.

DISCUSSION

To the best of our knowledge, no previous reports have assessed the effectiveness of pretransplant chest CT in predicting the development of TB after LT; thus, this study is the first to show that chest CT findings suggestive of healed TB can predict the development of TB in LT recipients. The number of patients enrolled in our study was large in comparison with previous studies, and LTBI screening or treatment was not performed before LT; this allowed an assessment of the natural course of TB after LT. Furthermore, most of the patients developed TB within 1 year of transplantation at the site of the healed TB. These findings suggest that most post-LT patients develop TB through the reactivation of old foci and not through new infections.

Because TSTs remain the standard method for detecting LTBIs, isoniazid is recommended for TST-positive organ transplant recipients.1, 4, 5 TSTs, however, cannot detect LTBIs in many LT recipients, and radiological features of a previous TB infection are a stronger risk factor for active posttransplant TB than positive TST results.11 Because TSTs have low sensitivity in immunocompromised patients and low specificity due to a previous bacillus Calmette-Guérin vaccination or an infection with nontuberculous mycobacteria, TSTs have several limitations in transplant recipients, especially in countries with an intermediate to high TB burden.13 Korea has an intermediate TB burden; thus, a treatment based on TST results in Korean LT recipients has not been determined. In addition, LT recipients are potentially at risk for hepatotoxicity associated with anti-TB treatments. Most transplant centers in Korea, including ours, neither test for LTBIs nor treat them in LT candidates.

The need for more accurate tests for detecting LTBIs in LT candidates has led to the use of interferon-γ release assays (IGRAs). The QuantiFERON-TB Gold assay (Cellestis, Ltd., Carnegie, Australia) has been found to be comparable to TSTs in the diagnosis of LTBIs,14 whereas the T-SPOT.TB assay (Oxford Immunotec, Abingdon, United Kingdom) has been found to be more useful for diagnosing LTBIs in these patients.15 Further studies of the best diagnostic methods for detecting LTBIs, including combinations of individual methods such as TSTs, IGRAs, and chest CT scans, are needed.

Many reports have described the usefulness of CT in finding additional cases of TB disease missed on CXRs under heterogeneous clinical conditions.16-18 Even theoretically, chest CT is better than a simple CXR for detecting calcified mediastinal and hilar lymph nodes. In our previous study,16 chest CT scans along with TSTs and IGRAs were effective in differentiating between subjects with active TB infections, subjects with LTBIs, and uninfected subjects in a contact investigation. Lee et al.17 described the usefulness of high-resolution CT scanning, which may be helpful in differentiating active TB from LTBIs in outbreak investigations of TB. Nakanishi et al.18 reported that high-resolution CT findings such as a tree-in-bud appearance, lobular consolidation, and large nodules were significantly associated with an increased risk for pulmonary TB. In this study, abnormal lesions suggestive of healed TB appeared more frequently in chest CT scans for the TB group versus the control group (40.0% versus 17.3%, respectively, P = 0.018). However, we should also recognize the limitations of CT scanning in the diagnosis or prediction of TB. Malaviya et al.19 reported that a chest CT scan was never positive unless a TST or an IGRA was positive in patients with inflammatory rheumatic disease who were treated with a tumor necrosis factor α inhibitor. In our study, the proportions of pre-LT abnormalities suggestive of healed TB did not significantly differ between the CT (40.0%) and CXR images (25.0%) for the TB group, and 15 patients (60%) who developed TB after transplantation had normal pre-LT chest CT findings. Collectively, these findings suggest that even though chest CT has the potential to play a role as a predictor of the future development of TB after LT and is possibly better than CXRs in this respect, chest CT alone has limited value as a predictor of the development of TB in LT recipients. Further studies that combine, for example, TSTs, IGRAs, and CXRs or CT are needed.

Traditionally, the development of TB in immunosuppressed patients occurs via the reactivation of latent foci and not via new infections, especially in areas with a low incidence of TB. Several recent studies, however, have suggested that TB recurs mostly via new infections rather than reactivation, especially in areas with a high incidence of TB.20, 21 Although our study was performed in a country with an intermediate TB burden, our results suggest that most of the patients developed TB via the reactivation of latent foci and not through new infections. That is, most of the patients who developed TB did so within 1 year after LT, with 70% developing TB at the site of the healed TB lesions.

This study has several limitations, including those inherent to all retrospective studies at a single center. In addition, not all patients underwent chest CT, and the medical records lacked information about the bacillus Calmette-Guérin vaccination status and detailed histories of previous TB infections and contact with TB-infected patients. Moreover, TSTs, the standard method for detecting LTBIs, were not evaluated in this study, primarily because TSTs have a high false-positive rate due to the high rates of bacillus Calmette-Guérin vaccination and a high false-negative rate due to immunosuppression.

In conclusion, our findings show that among those with post-LT TB, a pretransplant chest CT scan is more likely to show an LTBI than a CXR in a TB-endemic country. However, the usefulness of chest CT in predicting the development of TB after LT should be further investigated, especially in combination with the current standard method (TSTs) and/or new methods (IGRAs) for detecting LTBIs.

Ancillary

Advertisement