7
ORIGINAL ARTICLE Tuberculosis in Liver Transplant Recipients: Experience of a South Indian Liver Transplant Center Arikichenin Olithselvan, 1,2 Srinivas Rajagopala, 1 Mukul Vij, 1 Vivekanandan Shanmugam, 1 Naresh Shanmugam, 1 and Mohammed Rela 1 1 Department of Liver Transplantation and Hepatobiliary Surgery, Global Hospitals and Health City, Chennai, India; and 2 Department of Medicine, Jawaharlal Institute of Postgraduate Medical Education and Research, Puducherry, India Tuberculosis (TB) is a serious disease for liver transplant recipients (LTRs). Data on post–liver transplant TB from high-burden countries are scant. The aims of this study were to describe the prevalence of TB in LTRs from a high-prevalence area and to analyze the risk factors for the development of post–liver transplant TB. We performed a retrospective review of our database and a case-control study of identified cases with TB and age-matched LTRs without TB. The overall prevalence of TB in LTRs was comparable to the prevalence of TB in LTRs from low-prevalence countries (5/214 or 2.3%). A low rate of interferon-g release assay (IGRA) testing before liver transplantation was observed (68/214 or 31%). Most patients were screened clinically and with chest radiography alone before transplantation. TB developed variably after transplantation [median 5 72 days, inter- quartile range (IQR) 5 534 days]. The presentation was mostly extrapulmonary and/or disseminated (4/5 or 80%). When cases with posttransplant TB were compared with matched healthy LTRs, the presence of unexplained granulomas in explants (2/5 or 40%, P 5 0.01) was the only factor associated with the development of TB. When all explants showing granulomas were reviewed, TB (52.9%) remained the most common cause; however, in almost half (47.1%), other attributable causes were found. Patients were treated with a standard daily regimen for a median of 12 months (IQR 5 7.5 months). Posttransplant TB was asso- ciated with a high mortality rate (2/5 or 40%). In conclusion, we observed a low prevalence of TB in LTRs from a high- prevalence region. The presence of granulomas suggestive of TB in liver explants warrants isoniazid prophylaxis in the absence of disease. Post–liver transplant TB is associated with a high mortality rate. The roles of routine IGRA testing and isoniazid pro- phylaxis in a high-prevalence setting urgently need to be studied. Liver Transpl 20:960-966, 2014. V C 2014 AASLD. Received December 17, 2013; accepted April 21, 2014. Tuberculosis (TB) continues to be a global scourge. India has the highest TB burden and accounts for approximately one-fifth (21%) of the global disease burden. It is estimated that approximately 40% of the Indian population is infected with the TB bacillus, and the prevalence of smear-positive pulmonary cases remains approximately 75/100,000. 1 Liver transplantation is increasingly being per- formed for end-stage liver disease in India, with an estimated 1472 procedures across 22 centers by 2010. 2 There has been a significant improvement in survival after liver transplantation in India, with cen- ters reporting survival equivalent to survival in the West. 3 The huge burden of end-stage liver disease in Abbreviations: AFB, acid-fast bacillus; CMV, cytomegalovirus; CSF, cerebrospinal fluid; CTP, Child-Turcotte-Pugh; DDLT, deceased donor liver transplantation; HREL, isoniazid, rifampicin, ethambutol, and levofloxacin; HRZE, isoniazid, rifampicin, pyr- azinamide, and ethambutol; IGRA, interferon-g release assay; IQR, interquartile range; LDLT, living donor liver transplantation; LTBI, latent tuberculosis infection; LTR, liver transplant recipient; MELD, Model for End-Stage Liver Disease; MMF, mycopheno- late mofetil; NAFLD, nonalcoholic fatty liver disease; PCR, polymerase chain reaction; PJP, Pneumocystis jiroveci pneumonia; TB, tuberculosis; TST, tuberculin skin testing. The authors have no funding or conflicts of interest to declare. Address reprint requests to Arikichenin Olithselvan, M.D.,M.R.C.P., Department of Liver Transplantation and Hepatobiliary Surgery, Global Hos- pitals and Health City, Perumbakkam, Chennai, India 600100. Telephone: 191 044 22777777; FAX: 191 044 44777100; E-mail: [email protected] DOI 10.1002/lt.23903 View this article online at wileyonlinelibrary.com. LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases LIVER TRANSPLANTATION 20:960–966, 2014 V C 2014 American Association for the Study of Liver Diseases.

Tuberculosis in liver transplant recipients: Experience of a South Indian liver transplant center

Embed Size (px)

Citation preview

Page 1: Tuberculosis in liver transplant recipients: Experience of a South Indian liver transplant center

ORIGINAL ARTICLE

Tuberculosis in Liver Transplant Recipients:Experience of a South Indian Liver TransplantCenterArikichenin Olithselvan,1,2 Srinivas Rajagopala,1 Mukul Vij,1 Vivekanandan Shanmugam,1

Naresh Shanmugam,1 and Mohammed Rela1

1Department of Liver Transplantation and Hepatobiliary Surgery, Global Hospitals and Health City,Chennai, India; and 2Department of Medicine, Jawaharlal Institute of Postgraduate Medical Educationand Research, Puducherry, India

Tuberculosis (TB) is a serious disease for liver transplant recipients (LTRs). Data on post–liver transplant TB from high-burdencountries are scant. The aims of this study were to describe the prevalence of TB in LTRs from a high-prevalence area and toanalyze the risk factors for the development of post–liver transplant TB. We performed a retrospective review of our databaseand a case-control study of identified cases with TB and age-matched LTRs without TB. The overall prevalence of TB in LTRswas comparable to the prevalence of TB in LTRs from low-prevalence countries (5/214 or 2.3%). A low rate of interferon-grelease assay (IGRA) testing before liver transplantation was observed (68/214 or 31%). Most patients were screened clinicallyand with chest radiography alone before transplantation. TB developed variably after transplantation [median 5 72 days, inter-quartile range (IQR) 5 534 days]. The presentation was mostly extrapulmonary and/or disseminated (4/5 or 80%). When caseswith posttransplant TB were compared with matched healthy LTRs, the presence of unexplained granulomas in explants (2/5 or40%, P 5 0.01) was the only factor associated with the development of TB. When all explants showing granulomas werereviewed, TB (52.9%) remained the most common cause; however, in almost half (47.1%), other attributable causes were found.Patients were treated with a standard daily regimen for a median of 12 months (IQR 5 7.5 months). Posttransplant TB was asso-ciated with a high mortality rate (2/5 or 40%). In conclusion, we observed a low prevalence of TB in LTRs from a high-prevalence region. The presence of granulomas suggestive of TB in liver explants warrants isoniazid prophylaxis in the absenceof disease. Post–liver transplant TB is associated with a high mortality rate. The roles of routine IGRA testing and isoniazid pro-phylaxis in a high-prevalence setting urgently need to be studied. Liver Transpl 20:960-966, 2014. VC 2014 AASLD.

Received December 17, 2013; accepted April 21, 2014.

Tuberculosis (TB) continues to be a global scourge.India has the highest TB burden and accounts forapproximately one-fifth (21%) of the global diseaseburden. It is estimated that approximately 40% of theIndian population is infected with the TB bacillus,and the prevalence of smear-positive pulmonary casesremains approximately 75/100,000.1

Liver transplantation is increasingly being per-formed for end-stage liver disease in India, with anestimated 1472 procedures across 22 centers by2010.2 There has been a significant improvement insurvival after liver transplantation in India, with cen-ters reporting survival equivalent to survival in theWest.3 The huge burden of end-stage liver disease in

Abbreviations: AFB, acid-fast bacillus; CMV, cytomegalovirus; CSF, cerebrospinal fluid; CTP, Child-Turcotte-Pugh; DDLT,deceased donor liver transplantation; HREL, isoniazid, rifampicin, ethambutol, and levofloxacin; HRZE, isoniazid, rifampicin, pyr-azinamide, and ethambutol; IGRA, interferon-g release assay; IQR, interquartile range; LDLT, living donor liver transplantation;LTBI, latent tuberculosis infection; LTR, liver transplant recipient; MELD, Model for End-Stage Liver Disease; MMF, mycopheno-late mofetil; NAFLD, nonalcoholic fatty liver disease; PCR, polymerase chain reaction; PJP, Pneumocystis jiroveci pneumonia; TB,tuberculosis; TST, tuberculin skin testing.

The authors have no funding or conflicts of interest to declare.

Address reprint requests to Arikichenin Olithselvan, M.D., M.R.C.P., Department of Liver Transplantation and Hepatobiliary Surgery, Global Hos-pitals and Health City, Perumbakkam, Chennai, India 600100. Telephone: 191 044 22777777; FAX: 191 044 44777100; E-mail:[email protected]

DOI 10.1002/lt.23903View this article online at wileyonlinelibrary.com.LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

LIVER TRANSPLANTATION 20:960–966, 2014

VC 2014 American Association for the Study of Liver Diseases.

Page 2: Tuberculosis in liver transplant recipients: Experience of a South Indian liver transplant center

India, the increased use of living donor transplants,the increasing affordability to the population, and thecomparable success rates suggest that the number ofliver transplants is likely to increase even further inthe near future.3

TB is a significant complication for liver transplantrecipients (LTRs) and occurs in 0.47% to 2.3% ofrecipients worldwide.4,5 Recipients are vulnerable todeveloping TB through the reactivation of latenttuberculosis infections (LTBIs), reinfection in the com-munity, transmission from infected grafts, and noso-comial infections.5,6 Transplant recipients have a 36-to 74-fold higher risk of developing TB in comparisonwith the general population.5 The prevalence of TBamong transplant recipients is linked to the localprevalence of TB, which affects the baseline tubercu-lin positivity rate, the attendant risk of reactivation,and reinfection rates.7 The prevalence of post–renaltransplant TB varies from 13.3% in India8 to 15.2% inPakistan.9 Data on the prevalence of LTBIs and theprevalence of TB in LTRs are not available for Indiaand other high-burden countries. Routine tuberculinskin testing (TST) for LTBIs and isoniazid prophylaxisfor TST-positive LTRs are recommended,4 but theyare seldom implemented worldwide.6,10-12 Further-more, the role of tuberculin testing (TST), interferon-grelease assays (IGRAs), and isoniazid prophylaxis inLTRs remains undefined in high-prevalence areas,including India. This is due to the large baseline posi-tive rates for TST in high-prevalence areas and thepoor performance of TST in immunosuppressedpatients.13 Problems with the routine use of IGRAsinclude the indeterminate results of QuantiFERONtesting, the comparable performance of IGRA testingand TST in high-prevalence settings,14 and problemsin the interpretation of serial IGRA tests in high-transmission settings.15 Concerns about hepatotoxic-ity, utility in the face of ongoing high transmissionrates and lifelong immunosuppression, and the possi-bility of increasingly isoniazid-resistant TB with wide-spread isoniazid prophylaxis also complicatetreatment decisions for LTBIs in high-transmissionsettings.16 The risk factors for the development ofposttransplant TB also remain unclear. Liver trans-plantation has a significant interaction with the clini-cal presentation, diagnostic assays used for TB,treatment regimens (with several drug interactions),drug-related hepatotoxicity, and outcomes.6 Post-transplant TB remains a life-threatening disease withmortality rates ranging from 30% to 100% in variousseries.11,17 We reviewed our transplant database toaddress the prevalence of TB after liver transplanta-tion in a high-prevalence area and to identify the riskfactors for the development of TB in this setting.

PATIENTS AND METHODS

Study Location and Transplant Protocol

The study was conducted at Global Hospitals andHealth City (Chennai, India), which is a 750-bed terti-

ary hospital and liver transplant center. The livertransplant program was started in October 2009, anda computerized record of all recipients with theirfollow-up is maintained. All patients undergoing livertransplantation at our center undergo the same set oftests before transplantation. Subsequently, the initialfollow-up of 3 months is always at this center, andthe data are computerized after each visit. Ifthe patients are subsequently referred to a centerclose to their hometown for follow-up, this is super-vised by the hepatologist managing the patient atleast every 3 months. Patients with missing data arecategorized as lost to follow-up. The standard triple-immunosuppression regimen for LTRs consists oftacrolimus, mycophenolate mofetil (MMF), and ste-roids. Steroids are tapered and stopped approximately3 months after transplantation, and they are contin-ued over the long term only for patients with autoim-mune liver disease. All patients receive MMF (1 g/dayor 500 mg/day). Tacrolimus is dosed to maintain aserum level of 6 to 8 ng/mL until 3 months aftertransplantation and a serum level of 4 to 6 ng/mLafter that period. Acute cellular rejection is diagnosedon clinical grounds and is treated with increasingdoses of tacrolimus, MMF, and oral steroids. Nonres-ponders undergo liver biopsy and are then treatedwith an intravenous methylprednisolone bolusaccordingly. According to our protocol, all patientsreceive cytomegalovirus (CMV) prophylaxis from day 7to 3 months after transplantation (900 mg of valganci-clovir/day in divided doses). Pneumocystis jirovecipneumonia (PJP) prophylaxis is given only to high-risk individuals. These individuals have arbitrarilybeen defined as patients with diagnosed interstitiallung disease, chronic obstructive pulmonary disease,prior Pneumocystis carinii pneumonia, or the need tocontinue steroids more than 3 months after livertransplantation at our center. This decision is basedon the perceived low prevalence of PJP in our settingwith our immunosuppression protocol. Indeed, at thetime of this writing, no cases of PJP have been seen inmore than 200 transplants at our center. All prospec-tive recipients are screened by a pulmonologist for thepresence of active or treated TB before transplanta-tion, and a chest radiograph is reviewed for the pres-ence of active or healed TB. TST is not performedroutinely. Clinicians request QuantiFERON-TB Goldtesting [Cellestis, Ltd., St. Kilda, Australia] on thebasis of clinical assessments, and a pulmonologistmanaging a large load of TB patients assesses positiveresults. Patients receive isoniazid prophylaxis for 9months only if there is radiological evidence of healedTB, a past history of treated TB, definite contact witha smear-positive patient in the posttransplant period,or evidence of active or past TB in the donor. Nopatient has received isoniazid on the basis of theresults of the QuantiFERON-TB GOLD test. For livingdonors, the screening for TB is similar to that usedfor prospective recipients; donors are screened clini-cally and with chest radiography. For deceaseddonors, this is supplemented with chest and abdomen

LIVER TRANSPLANTATION, Vol. 20, No. 8, 2014 OLITHSELVAN ET AL. 961

Page 3: Tuberculosis in liver transplant recipients: Experience of a South Indian liver transplant center

computed tomography and with the evaluation of anypathology suspicious for TB.

Study Plan and Design

A retrospective analysis of the liver transplant databasewas performed. All patients in the database were identi-fied, and the period of follow-up was calculated inperson-years. The follow-up duration was censored tothe last follow-up visit (clinical or telephonic) or death asappropriate. All patients diagnosed with TB in the post-transplant period were identified. Patients were definedas having active TB after liver transplantation if (1) apositive culture for Mycobacterium tuberculosis or a posi-tive polymerase chain reaction (PCR) test forM. tuberculosis DNA was obtained from a specimenrelated to the patient’s signs or symptoms or (2) acid-fast bacilli (AFBs), a caseating granuloma, or both weredemonstrated by a histopathological examination alongwith a clinical response to anti-TB therapy. Dissemi-nated TB was diagnosed if 2 or more noncontiguoussites showed evidence of TB. The medical records of theidentified cases were retrieved, and the data wereabstracted in a standardized manner. All age- and sex-matched controls who had follow-up for at least 1 yearwere noted from our computerized database and thenentered into a spreadsheet. One hundred of 214 patientscould be matched by age (65 years) and sex. To restrictthe number of controls, we decided on a ratio notexceeding 1:8; 40 controls were randomly selected fromthe list of matched controls before the data analysis withthe randomization software available (SPSS 14). Data onrisk factors for TB were abstracted for both cases andcontrols with a predefined pro forma. The data includedthe following: sex, age, indication for liver transplanta-tion, pretransplant Model for End-Stage Liver Disease(MELD) score, Child score, coexisting diabetes mellitus,requirement for renal support before transplantation,and CMV status. Abstracted postoperative variablesincluded the following: details of the immunosuppres-sion regimen, rejection episodes, use of bolus methyl-prednisolone, additional immunosuppression (OKT3 andbasiliximab), posttransplant CMV infections, clinical pre-sentation of TB, diagnosis of TB, and outcomes.

Statistics and Ethics

Continuous variables are presented in a descriptivefashion [means and standard deviations or mediansand interquartile ranges (IQRs)], and categorical vari-ables are presented as frequencies and proportions.Continuous variables were compared with the Mann-Whitney U test, and categorical variables were com-pared with Fisher’s exact test. Statistical significancewas assessed at the 2-sided P�0.05 level. The insti-tutional ethics board approved the study.

RESULTS

In all, 271 LTRs were identified since the transplantprogram began at our center. Two hundred fourteen

TA

BL

E1

.D

escri

pti

on

of

the

Cli

nic

al

Ch

ara

cte

risti

cs

of

LT

Rs

Wit

hP

ostt

ran

sp

lan

tT

B

Case

Age

(Yea

rs)/

Sex

Eti

olo

gy

of

Cir

rhosi

s

Pre

tran

spla

nt

Sev

erit

y

Type

of

Tra

nsp

lan

tC

om

orb

idit

ies

Tim

eFro

m

Tra

nsp

lan

t

(Days)

Pre

sen

tati

on

Dia

gn

osi

sTh

erapy

Ou

tcom

e

ME

LD

Sco

re

CTP

Sco

re

164/m

ale

NA

FLD

19

11

LD

LT

Dia

bet

es,

hyper

ten

sion

60

Men

ingit

isC

SF

AFB

1,

TB

PC

R1

HR

ZE

for

18

mon

ths

Alive

,re

curr

ent

tuber

culo

ma

236/m

ale

Cry

pto

gen

ic20

11

DD

LT

Non

e150

Tu

ber

culo

ma,

skin

lesi

on

sS

kin

bio

psy

AFB

1

HR

ZE

for

12

mon

ths

Alive

,w

ell

357/fe

male

Hep

ati

tis

C13

10

LD

LT

Non

e42

Ple

ura

lef

fusi

on

,asc

ites

Ple

ura

lflu

id,

asc

ites

AFB

1

HR

ZE

for

14

mon

ths

Die

d

454/m

ale

NA

FLD

>25

12

LD

LT

Dia

bet

es,

hyper

ten

sion

1020

Dis

sem

inate

dm

ilia

ryTB

,Fou

rnie

rgan

gre

ne

Skin

bio

psy

AFB

1

HR

ZE

for

5days

Die

d

561/m

ale

NA

FLD

15

9D

DLT

Dia

bet

es,

hyper

ten

sion

72

Fev

er,

ple

ura

lef

fusi

on

Ple

ura

lflu

idTB

PC

R1

On

goin

gH

RE

LA

live

,w

ell

962 OLITHSELVAN ET AL. LIVER TRANSPLANTATION, August 2014

Page 4: Tuberculosis in liver transplant recipients: Experience of a South Indian liver transplant center

of these patients (78.9%) had a follow-up of at least ayear and were selected. The median follow-up for thiscohort was 1.5 person-years (IQR 5 1.8 person-years).The vast majority of the recipients were adults (165/214 or 71.9%) and Indians (148/214 or 69%) andunderwent living related transplantation (141/214 or65.7%). All recipients underwent ABO-compatibletransplantation. Five of the 214 patients (2.3%) devel-oped TB after transplantation. The median time to thedevelopment of TB was 72 days (IQR 5 534 days) afterliver transplantation (Table 1). All cases of posttrans-plant TB with explant granulomas occurred within thefirst 40 transplant recipients. The clinical details ofthe individual cases are provided in Table 1.

The presentation was mostly extrapulmonary (4/5or 80%) and disseminated (3/5 or 75%). The diagno-sis was made with a positive Ziehl-Neelsen smear forAFBs in concurrence with M. tuberculosis–specificPCR for all patients, often from sites that are typicallypaucibacillary in immunocompetent patients [skinbiopsy, 2/5 or 40%; pleural and/or ascitic fluid, 2/5or 40%; and cerebrospinal fluid (CSF), 1/5 or 20%].Most patients were treated with a standard anti-TBregimen [isoniazid, 5 mg/kg; rifampicin, 10 mg/kg;pyrazinamide, 25 mg/kg; and ethambutol, 20 mg/kg),and tacrolimus levels were adjusted to compensate forthe interaction with rifampicin. Patients were treatedfor a median of 12 months (IQR 5 7.5 months). Drug-related hepatotoxicity (0/5 or 0%) and cellular rejec-tion (0/5 or 0%) did not occur for any of the 5patients. Four patients (4/5 or 80%) completed thetreatment with no complications; treatment for 1patient was interrupted by an early death. TB wasassociated with high mortality (2/5 or 40%) and mor-bidity rates (recurrent tuberculomas with raised intra-

cranial pressure, 1/5 or 20%; recurrent low serum-ascites albumin gradient ascites, 1/5 or 20%)

The etiology of liver disease, pretransplant diseaseseverity, type of transplant, nationality, immunosuppres-sion regimen, administration of additional immunosup-pression (OKT3, basiliximab, and methylprednisolone),and concurrent renal failure had no influence on thedevelopment of posttransplant TB (Table 2). The pres-ence of granulomas suggestive of TB in the explant liverwas associated with the development of TB (P5 0.010).

The results of the review of our transplant explantdatabase and the examination of the prevalence ofgranulomas in explants and their significance are pre-sented in Table 3. Explants showing granulomas werefound for 6.27% of the LTRs (17/271). The indexpatients for these explants had a median follow-up of11 months (IQR 5 14.50 months) after transplanta-tion. The procedure (transarterial chemoembolization),drug history (interferon, sorafenib, steroids, and basi-liximab), or both relevant to the development of gran-ulomas was noted in 29.4% (5/17). Granulomas wereportal or lobular in 76.4%, and caseation was seen in23.5% (4/17). AFBs by the Ziehl-Neelsen techniquewere universally negative. When a combined clinico-pathological meeting made a final diagnosis for theetiology of explant granulomas, an alternative causeof TB was determined for approximately half (8/17 or47.05%). Posttransplant TB developed in 2 of 17patients (11.8%) with explant granulomas suggestiveof TB.

DISCUSSION

Data from renal transplantation suggest a high LTBIburden in India,8,18 and the prevalence of post–renal

TABLE 2. Comparison of Clinical Characteristics and Risk Factors for LTRs With TB and Controls (Recipients)

Without TB

Characteristic

Posttransplant

LTRs With

TB (n 5 5)

Controls

(n 5 40) P Value

Age (years)* 54.4 6 10.9 42.67 6 22.0 0.22†

Sex: male/female (n/n) 4/1 33/7 1.00Etiology: alcohol versus non–alcohol-related (n/n) 0/5 9/31 0.57Pretransplant MELD score* 18.40 6 4.67 18.74 6 5.51 0.98†

Pretransplant CTP score* 10.60 6 1.14 9.36 6 1.65 0.10†

Transplant type: LDLT/DDLT (n/n) 3/2 26/14 1.00Pretransplant renal failure: yes/no (n/n) 1/4 9/31 1.00Nationality: Indian/foreign (n/n) 5/0 27/13 0.30Anti-thymocyte globulin administration: yes/no (n/n) 0/5 1/39 1.000Basiliximab administration: yes/no (n/n) 0/5 7/33 0.58Rejection episodes treated with methylprednisolone: yes/no (n/n) 0/5 5/35 0.46Postoperative isoniazid: yes/no (n/n) 0/5 1/39 1.000Explants showing granulomas: yes/no (n/n) 2/3 0/40 0.010Posttransplant renal failure: yes/no (n/n) 1/4 0/40 0.11

*The data are presented as means and standard deviations.†Mann-Whitney U test (Fisher’s exact test when not otherwise specified).

LIVER TRANSPLANTATION, Vol. 20, No. 8, 2014 OLITHSELVAN ET AL. 963

Page 5: Tuberculosis in liver transplant recipients: Experience of a South Indian liver transplant center

transplant TB varies from 13.3% in India8 to 15.2% inPakistan.9 The prevalence of LTBIs in LTRs in India isunknown; to the best of our knowledge, this is thefirst study of LTRs from a high-prevalence area. Ourstudy found a low prevalence of TB in this settingcomparable to data from a low-prevalence area. Possi-ble explanations include the lower intensity of immu-nosuppression (in comparison with renal transplants),a selection bias for recipients with respect to theirbaseline LTBI status, and the relatively short durationof the follow-up. However, most cases of TB aftertransplantation develop within the first year5; we

believe that our results reflect a truly low prevalenceof TB in this setting, and this is in concordance withself-assessments from other liver transplant centersin India (personal communication). More data fromother large centers in high-prevalence regions andlong-term follow-up of our cohort will be needed toconfirm this.

Tuberculin testing is not practiced at our center forseveral reasons; these include the poor performanceof tuberculin in patients with chronic liver disease19

and a lack of information on managing patients withpositive TST results16 and a universal national TBvaccination program (bacillus Calmette-Guerin). Aresponse to TST may signify an infection withM. tuberculosis, an infection with nontubercular Myco-bacteria, or, infrequently, bacillus Calmette-Guerinvaccination. Furthermore, in a high-prevalence set-ting, a positive TST result does not imply a recentconversion with its attendant implications for isonia-zid prophylaxis, and the risk of reactivation falls expo-nentially with each year of TST conversion. Data onTST testing before renal transplantation in India sug-gest a very low positive predictive value of 5.2% to23.8% for the development of posttransplant TB.20

The utility of IGRAs in this setting is also limited by alack of additional information about TST versus IGRAtesting14; other problems include the interpretation ofserial IGRA results21 and persistent elevations inM. tuberculosis–induced interferon-g levels in the set-ting of high transmission rates.15 Clinicians at our

TABLE 3. Clinical Characteristics of All LTRs With

Explants Showing Granulomas (17/271 or 6.27%) and All

LTRs Receiving Isoniazid (7/174)

Characteristic Value

Age (years)* 45.47 6 19.52Sex: male/female (n/n) 13/4Etiology of cirrhosis leading toliver transplant [n (%)]†

NAFLD 4/17 (23.5)Hepatitis C 5/17 (29.4)Autoimmune‡ 3/17 (17.6)Hepatocellular carcinoma 4/17 (23.5)Cryptogenic 2/17 (11.8)Other§ 2/17 (11.8)

Relevant drug/procedure history[n (%)]k

5/17 (29.4)

Follow-up since livertransplantation (months)¶

11 (14.50)

QuantiFERON-TB Gold assay[n/N (%)]

Available 4/17 (23.5)Positive 2/17 (11.8)

GranulomasExtent [n/N (%)]

Occasional 9/17 (52.94)Extensive 8/17 (47.06)

Distribution [n/N (%)]Portal 6/17 (35.29)Lobular 7/17 (41.18)Both 3/17 (17.65)Interface 1/17 (5.9)

Caseation [n/N (%)] 4/17 (23.5)Positivity for AFB and/or

M. tuberculosis [n/N (%)]0/17 (0)

Clinicopathological etiology[n/N (%)]†

TB 9/17 (52.94)Drug-related 2/17 (11.8)Foreign body 2/17 (11.8)Sarcoidosis 1/17 (5.9)Sclerosing cholangitis 1/17 (5.9)Tumor-related 1/17 (5.9)Unknown 2/17 (11.8)

Change in etiology for trans-plant after explant histopathology[n/N (%)]

1/17 (5.9)

Outcome [n/N (%)]Alive 13/17 (76.47)Died 4/17 (23.53)

TABLE 3. Continued

Characteristic Value

TB after liver transplantation[n/N (%)]

2/17 (11.8)

Isoniazid administrationIndication (n/N)

Donor TB 2/7Previously treated TB 3/7Granulomas in explanted liver 2/7

Drug-related hepatitis [n/N(%)]¶

0/7 (0)

TB after isoniazid prophylaxis[n/N (%)]

0/6 (0)

*The data are presented as means and standarddeviations.†More than 1 etiology was possible for some patients.‡Autoimmune cirrhosis (1/17 or 5.9%), primary biliarycirrhosis/autoimmune overlap (1/17 or 5.9%), andulcerative colitis/primary sclerosing cholangitis (1/17 or5.9%).§Tyrosinemia leading to hepatocellular carcinoma (1/17or 5.9%) and primary hyperoxaluria (1/17 or 5.9%).kTransarterial chemoembolization (2/17), interferon(2/17), sorafenib (1/17), steroids (1/17), and basiliximab(1/17).¶The data are presented as medians and IQRs.#Six of 7 patients completed 9 months of therapy.

964 OLITHSELVAN ET AL. LIVER TRANSPLANTATION, August 2014

Page 6: Tuberculosis in liver transplant recipients: Experience of a South Indian liver transplant center

center use QuantiFERON-TB Gold (Cellestis) for LTRsunder clinical suspicion. We identified a total of 68tests (68/210 or 31%) in this setting; 56 were nega-tive, 4 were intermediate, and another 8 were positive.None of the 8 positive patients developed TB after aminimum follow-up of 12 months. The infrequent useof the test reflects concerns about the lack of utility ofthe QuantiFERON-TB Gold test and the costs involvedin IGRA testing. Our frequency of testing is compara-ble to that found by the Group for the Study of Infec-tion in Transplant Recipients: only 31% of thetransplant recipients (297/926) had undergone TSTdespite guidelines recommending LTBI testing for allpatients undergoing liver transplantation.12,22 Ourdata on the prevalence of TB suggest that the currentclinical approach to screening patients and offeringisoniazid on the basis of clinical indications ratherthan tuberculin testing may be appropriate in a high-prevalence area.

Our posttransplant TB cohort is too small for anyfirm conclusions on associations with the develop-ment of TB to be derived. We did find an associationbetween granulomas in the explant liver and post-transplant TB. A wide variety of disease processescause hepatic granulomas.23 In patients with sponta-neously healed TB, healed granulomas may persist intissues, and M. tuberculosis has been cultured fromcalcified granulomas. During the initial period of ourtransplant program (40 liver transplants), the signifi-cance and management of these healed granulomasin the explanted liver were debated because no clearguidelines were available. Clinical equipoise existedbecause potentially up to 40% to 70% of Indians werelikely to have healed occasional granulomas in theirtissues, infection was often remote, immunosuppres-sion after transplantation was lifelong (whereas isoni-azid prophylaxis lasted 9 months), and there werehigh rates of reported isoniazid resistance by M. tuber-culosis.24 When 2 of these index patients developedposttransplant TB, we re-evaluated our protocol toinitiate isoniazid prophylaxis for 9 months at 6 weeksafter transplantation for all patients with granulomasin the explanted liver without an alternative cause.This was thought to be adequate to reduce the risk ofreactivation during the period when the intensity ofimmunosuppression was highest. Notably, in approxi-mately half (47.1%), other causes for granulomaswere found, and this stresses the importance of rou-tine clinicopathological reviews of explant pathology.Details for the index patients, explant granulomatouspathology, and isoniazid prophylaxis are provided inTable 3.

The present study addresses the paucity of data forpost–liver transplant TB in a high-prevalence area andincludes a complete analysis of all transplant patientswith at least 1 year of follow-up at our center, uniquedata on the practice and problems of screening beforeliver transplantation in a high-prevalence area, anddetails on explanted liver pathology relevant to further-ing our understanding of post–liver transplant TB.However, the small number of TB cases limits the sta-

tistical analysis of associations of other clinical varia-bles and the development of TB. Long-term follow-upof our expanding cohort and multicenter data fromhigh-prevalence areas will clarify this.

In conclusion, we have reported a low prevalence ofTB after liver transplantation in a high-prevalencearea. The presence of explant granulomas in theabsence of an alternative cause is a risk factor for thedevelopment of posttransplant TB, and isoniazid pro-phylaxis should be offered to all such patients. Mostpatients with posttransplant TB can be managed safelywith a standard TB regimen with adjustments of calci-neurin inhibitor levels. Post–liver transplant TB isassociated with high mortality and morbidity rates.Studies on pretransplant tuberculin and IGRA testingand isoniazid prophylaxis urgently need to be per-formed in a high-prevalence area.

REFERENCES

1. World Health Organization. Tuberculosis (TB). http://www.who.int/tb/data. Accessed April 2014.

2. Vibha Varma NM, Nundy S. Liver transplantation inIndia—the history. In: Nundy S, ed. Liver Transplanta-tion. India: Elsevier; 2011:8-14.

3. Kakodkar R, Soin A, Nundy S. Liver transplantation inIndia: its evolution, problems and the way forward. NatlMed J India 2007;20:53-56.

4. Holty JE, Gould MK, Meinke L, Keeffe EB, Ruoss SJ.Tuberculosis in liver transplant recipients: a systematicreview and meta-analysis of individual patient data. LiverTranspl 2009;15:894-906.

5. Singh N, Paterson DL. Mycobacterium tuberculosis infec-tion in solid-organ transplant recipients: impact andimplications for management. Clin Infect Dis 1998;27:1266-1277.

6. Clemente WT, Faria LC, Lima SS, Vilela EG, Lima AS,Velloso LF, et al. Tuberculosis in liver transplant recipi-ents: a single Brazilian center experience. Transplanta-tion 2009;87:397-401.

7. Subramanian A, Dorman S; for AST Infectious DiseasesCommunity of Practice. Mycobacterium tuberculosis insolid organ transplant recipients. Am J Transplant 2009;9(suppl 4):S57-S62.

8. John GT, Shankar V, Abraham AM, Mukundan U,Thomas PP, Jacob CK. Risk factors for post-transplanttuberculosis. Kidney Int 2001;60:1148-1153.

9. Naqvi A, Rizvi A, Hussain Z, Hafeez S, Hashmi A, AkhtarF, et al. Developing world perspective of posttransplanttuberculosis: morbidity, mortality, and cost implications.Transplant Proc 2001;33:1787-1788.

10. Benito N, Sued O, Moreno A, Horcajada JP, Gonz�alez J,Navasa M, Rimola A. Diagnosis and treatment of latenttuberculosis infection in liver transplant recipients in anendemic area. Transplantation 2002;74:1381-1386.

11. Hsu MS, Wang JL, Ko WJ, Lee PH, Chou NK, Wang SS,et al. Clinical features and outcome of tuberculosis insolid organ transplant recipients. Am J Med Sci 2007;334:106-110.

12. Aguado JM, Herrero JA, Gavald�a J, Torre-Cisneros J,Blanes M, Ruf�ı G, et al. Clinical presentation and out-come of tuberculosis in kidney, liver, and heart trans-plant recipients in Spain. Spanish TransplantationInfection Study Group, GESITRA. Transplantation 1997;63:1278-1286.

13. Lee SS, Chou KJ, Su IJ, Chen YS, Fang HC, Huang TS,et al. High prevalence of latent tuberculosis infection in

LIVER TRANSPLANTATION, Vol. 20, No. 8, 2014 OLITHSELVAN ET AL. 965

Page 7: Tuberculosis in liver transplant recipients: Experience of a South Indian liver transplant center

patients in end-stage renal disease on hemodialysis:comparison of QuantiFERON-TB GOLD, ELISPOT, andtuberculin skin test. Infection 2009;37:96-102.

14. Pai M, Gokhale K, Joshi R, Dogra S, Kalantri S,Mendiratta DK, et al. Mycobacterium tuberculosis infec-tion in health care workers in rural India: comparison ofa whole-blood interferon gamma assay with tuberculinskin testing. JAMA 2005;293:2746-2755.

15. Pai M, Joshi R, Dogra S, Mendiratta DK, Narang P, DhedaK, Kalantri S. Persistently elevated T cell interferon-gamma responses after treatment for latent tuberculosisinfection among health care workers in India: a prelimi-nary report. J Occup Med Toxicol 2006;1:7.

16. Agarwal R. Treatment of latent tuberculous infection inIndia: is it worth the salt? Lung India 2005;22:105-106.

17. Zhang XF, Lv Y, Xue WJ, Wang B, Liu C, Tian PX, et al.Mycobacterium tuberculosis infection in solid organtransplant recipients: experience from a single center inChina. Transplant Proc 2008;40:1382-1385.

18. Sakhuja V, Jha V, Varma PP, Joshi K, Chugh KS. Thehigh incidence of tuberculosis among renal transplantrecipients in India. Transplantation 1996;61:211-215.

19. Celikbilek M, Selcuk H, Yilmaz U. The effect of hepato-tropic virus (HBV-HCV) infections on tuberculin skin test

in patients with cirrhosis. Turk J Gastroenterol 2012;23:234-238.

20. Agarwal SK, Gupta S, Bhowmik D, Mahajan S. Tubercu-lin skin test for the diagnosis of latent tuberculosis dur-ing renal replacement therapy in an endemic area: asingle center study. Indian J Nephrol 2010;20:132-136.

21. Zwerling A, Benedetti A, Cojocariu M, McIntosh F,Pietrangelo F, Behr MA, et al. Repeat IGRA testing inCanadian health workers: conversions or unexplainedvariability? PLoS One 2013;8:e54748.

22. Aguado JM, Torre-Cisneros J, Fort�un J, Benito N, MeijeY, Doblas A, Mu~noz P. Tuberculosis in solid-organ trans-plant recipients: consensus statement of the Group forthe Study of Infection in Transplant Recipients (GESI-TRA) of the Spanish Society of Infectious Diseases andClinical Microbiology. Clin Infect Dis 2009;48:1276-1284.

23. Bhardwaj SS, Saxena R, Kwo PY. Granulomatous liverdisease. Curr Gastroenterol Rep 2009;11:42-49.

24. Deepa D, Achanta S, Jaju J, Rao K, Samyukta R,Claassens M, et al. The impact of isoniazid resistance onthe treatment outcomes of smear positive re-treatmenttuberculosis patients in the state of Andhra Pradesh,India. PLoS One 2013;8:e76189.

966 OLITHSELVAN ET AL. LIVER TRANSPLANTATION, August 2014