mcilleron tuberculosis 2007

7/24/2019 McIlleron Tuberculosis 2007

1/9

This article was published in an Elsevier journal. The attached copyis furnished to the author for non-commercial research and

education use, including for instruction at the authors institution,sharing with colleagues and providing to institution administration.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elseviers archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright
http://www.elsevier.com/copyrighthttp://www.elsevier.com/copyright


2/9

Author's personal copy

Available at www.sciencedirect.com

ournal homepage: http://intl.elsevierhealth.com/journals/tube

Rifampin levels, interferon-gamma release andoutcome in complicated pulmonary tuberculosis

Helen McIlleron a, , Marcia L. Watkins b , Peter I. Folb a ,Stanley R. Ress b , Robert J. Wilkinson c,d

a Department of Medicine, Division of Clinical Pharmacology, University of Cape Town, K-45 Old Main Buliding,Groote Schuur Hospital, Observatory, Cape Town, Western Cape, 7925, South Africab Department of Medicine, Division of Clinical Immunology, University of Cape Town, Groote Schuur Hospital, South AfricacDepartment of Medicine, Institute of Infectious Diseases and Molecular Medicine, Faculty of Health Sciences,University of Cape Town, South Africad Wellcome Trust Center for Research in Clinical Tropical Medicine, Imperial College London, Division of Medicine,Wright Fleming Institute, London, UK

Received 14 May 2007; received in revised form 27 July 2007; accepted 10 August 2007

KEYWORDSMycobacteriumtuberculosis ;Cytokines;Pharmacokinetics

SummaryFactors that relate to medium-term outcome in patients with pulmonary tuberculosis(PTB) who have completed the 2-month intensive phase of treatment are incompletelyunderstood. The relationship between in vitro production of interferon-gamma (IFN- g ),interleukins (ILs)-5 and -10 and drug levels determined after 2 months of drug therapy, tooutcome at 24 months was studied prospectively. Cytokine concentrations weredetermined from culture supernatants after stimulation of whole blood with puriedprotein derivative (PPD) of Mycobacterium tuberculosis . Plasma concentrations of rifampin, isoniazid, pyrazinamide and ethambutol were determined by high-performanceliquid chromatography. The treatment failure and relapse free survival probability was0.54 (95% CI: 0.400.67) at 24 months. In multivariate analysis of parameters at 2 monthsthe strongest positive associations with disease free survival were IFN- g response to PPD

( p 0.002) and serum creatinine ( p 0.001). Drug concentrations were not associatedwith outcome although rifampin exposure correlated with IFN- g response to PPD( p 0.0132). These data suggest that the ability to mount a recall immune response toM. tuberculosis may inuence treatment outcome. The data support the idea to identifypersons at risk of a poor treatment outcome by monitoring of the in vitro response totuberculosis antigens.& 2007 Elsevier Ltd. All rights reserved.

ARTICLE IN PRESS

1472-9792/$ - see front matter & 2007 Elsevier Ltd. All rights reserved.doi: 10.1016/j.tube.2007.08.002

Corresponding author. Tel.: +27 21 406 6292; fax: +27 21 448 1989.E-mail addresses: [email protected] (H. McIlleron) , [email protected] (M.L. Watkins) , [email protected] (P.I. Folb) ,

[email protected] (S.R. Ress) , [email protected] (R.J. Wilkinson) .

Tuberculosis (2007) 87 , 557564


3/9


Introduction

Tuberculosis is a globally important cause of mortality. Theprincipal control measure is detection and treatment of infectious cases of pulmonary tuberculosis (PTB). Rifampin-

based multi-drug regimens administered as part of a directlyobserved therapy (DOTS) strategy to ambulatory patientsare cost effective albeit with the need to be administeredfor at least 6 months. The cure rate determined by sputumclearance is a key indicator of the performance of treatmentprogrammes.

However, some subgroups do badly and require hospitalisa-tion at least for the initial 4-drug phase of therapy. Whilst theincreasing incidence of drug resistant disease is a majorconcern, 1 drug sensitive disease also presents treatmentdifculties requiring inpatient management. Unemployment,overcrowded living conditions, smoking, extensive or recur-rent disease, underlying lung disease, and advanced age orimmunosuppression can contribute. 25 The bioavailability of drugs during therapy is important but not routinely ascer-tained in patients. In these circumstances it is useful to knowwhether therapy is effective such that the patient may bedischarged to ambulant DOTS care after completing 2 monthsof treatment. This assessment is classically based uponsputum smear conversion. However, the latter is not alwaysaccompanied by culture conversion, the results of which maynot become available until many weeks after the patient hasleft the inpatient facility and possibly lost to follow-up. Arapid method to predict poor prognosis could contribute toincreased vigilance thus improving cure rates.

A variety of immune parameters, including antibody,cellular proliferative, cytokine, and acute phase responses,change during successful treatment of tuberculosis. 69

Whether the immune response thereby contributes to thesuccess of therapy is conjectural: an equal cure rate foruncomplicated tuberculosis is observed in HIV-infectedpersons, although the subsequent relapse rate is higher. 10

The extent to which the changes represent phenotypicchange in the immune response or recirculation of cells fromdisease sites is also unclear. 8,11 There is considerableinterest in the possibility of assisting diagnosis and monitor-ing therapy via in vitro detection of interferon-gamma(IFN-g ) released in response to Mycobacterium tuberculosis antigens. 11 13

We evaluated factors that may inuence the mediumterm outcome of PTB in a cohort of 142 patients requiringhospitalisation for drug sensitive PTB in an area of SouthAfrica with a very high incidence of tuberculosis. We relatedclinical, immunological and pharmacological determinationsat the 2-month time point to outcome at 24 months. Thestrongest predictor of outcome was the level of IFN- greleased into culture in response to puried proteinderivative (PPD) of M. tuberculosis . This level in turncorrelated with the simultaneously determined area underthe curve (AUC) for rifampin, but no other antituberculardrug. The signicance of, and practical possibilities raisedby, these observations are discussed.

Materials and methods

The study was approved by the University of Cape TownResearch Ethics Committee (REC 122/99), was conducted inaccordance with the Helsinki Declaration of 1975 as revisedin 1983 and was performed in collaboration with theregional health authorities. Participants provided writtenconsent for inclusion. Patients ( n 142) admitted to theregional tuberculosis hospital with PTB 2 months previouslywere sequentially enrolled ( Figure 1). They were referred tohospital for reasons including poor response to treatment,suspected non-adherence, debility, severe or complicateddisease and poor socioeconomic circumstances. Patientsfrom whom M. tuberculosis resistant to rifampin or isoniazidhad been isolated were excluded. Daily antituberculartreatment doses based on body weight were directlyobserved by hospital staff. Determinations at the 2 monthtime point included demographic and treatment factors, HIVstatus (AxSYM HIV Ag/Ab Combo, Abbott Diagnostics,Germany), cytokine responses to PPD, lymphocyte subsetcounts, serum albumin, serum creatinine and plasmaconcentrations of rifampicin, isoniazid, pyrazinamide andethambutol. Following discharge after 2 months of treat-ment in hospital, participants were actively followed-upthrough the regional clinics using 3- and then 6-monthlysputum smear microscopy and culture of a single sputumspecimen, and clinic records, until 24 months after theiradmission to the hospital.

ARTICLE IN PRESS

142 patients

cytokine levels were evaluated in 112 patients

12 died

4 failedtreatment

4 died early( 6 months)

8 died later (7 to 24months)

12 relapsed

22 had agood

outcome at24

months

62 lost tofollow up

hiv-infection 0 1 3 3 3 4retreatment 4 4 5 6 13 39

male 2 2 1 7 7 31

Figure 1 Summary of recruitment and of cohort outcome. Patients with disease relapse had a positive sputum by microscopy orculture after treatment completion. Patients with a good clinical outcome at 24 months were dened as those without treatmentfailure, relapse or death and who were not lost to follow-up at the end of the observation period.

H. McIlleron et al.558


4/9


Blood assays

The steady-state plasma concentrations of rifampin, iso-niazid, pyrazinamide and ethambutol were determined over8h to generate the AUC 0 8 as previously described. 14 Dual

parameter ow cytometric analysis (performed on an EpicsProle II [Coulter], using CD3RD1 and either CD4-FITC orCD8-FITC markers [Coulter, Immunology, Hialeah, Florida])and PPD stimulated cytokine levels were determined in 112patients. To determine the cytokine response, whole blooddiluted at a nal concentration of 1:5 in RPMI 1640(BioWittaker, Cambrex) was cultured with PPD fromM. tuberculosis (Central Veterinary Laboratory, Weybridge,UK) at a nal concentration of 3 mg/ml for 5 days at 37 1C in ahumidied atmosphere containing 5% CO 2 . Supernatantconcentrations of IFN- g , interleukin (IL)-5 and IL-10 weredetermined by ELISA using commercially available antibodypairs (BD PharMingen, San Diego, CA). The lower limit of detection of the assays was 8 pg/ml; computed levels belowthis were assigned 8 pg/ml. The PPD stimulated incrementswere calculated by deduction of the levels detected in theunstimulated blood samples. If the result was negative, itassumed a value of 0 pg/ml. A positive control for the wholeblood culture was provided by phytohaemaglutinin (PHA:Wellcome Research Laboratories, Beckenham, UK) at a nalconcentration of 5.75 10 3 mitogenic units/ml.

Analysis

Stata version 8.2 (Stata Corp., College Station, TX 77845)was used to compute summary statistics and for statisticalmodelling. Kaplan Meier survival analysis described tuber-

culosis-free survival from 2 until 24 months after hospitaladmission using time to the rst positive sputum smear orculture (in cases diagnosed with treatment failure orrelapse), or death. Observations were censored if the patientwas alive and not sputum-positive at the time of loss tofollow-up. Cox proportional hazards models were used toreport associations of the covariate factors (age, sex, newor retreatment category, HIV status, cytokine responses toPPD, lymphocyte subset counts, drug concentrations andserum albumin and creatinine as biochemical markers of nutrition and disease severity) with the 24-month outcome.For the multivariate analysis, variable selection was initiallyby an automated backwards stepwise process (variables with pX 0.055 being removed from the model, and retained if po 0.050), followed by a forwards stepwise procedure basedon the contribution of individual variables to the overall t of the model. The model assumptions of proportionality and theappropriate form of the covariates in the model werechecked using Schoenfeld residuals distributions.

Results

Recruitment and clinical outcome data are summarised inFigure 1. Logistical constraints prevented cytokine evalua-tion in 30 patients. The evaluated patients tended to beslightly older than those not included in the analysis (median36 years, IQR: 2946 vs. 31.5 years, IQR: 26 40; p 0.08),and 14 HIV-infected patients were evaluated compared tonone in the group not evaluated. The proportions of

retreatment and male patients were similar between theevaluated and unevaluated groups (63% vs. 67% retreatedand 45% vs. 47% male, respectively); likewise albumin andcreatinine concentrations were closely similar between thetwo groups (mean 34.1 g/l, SD: 6.0 vs. 33.0g/l, SD: 4.7 foralbumin, and median 74 mmol/l, IQR: 65 83 vs. median73.5 mmol/l, IQR: 65 84). Sixty-two observations (55%) werecensored as the patients were alive and did not have eitherpositive sputum microscopy or culture results at the time of their last study visit. The in study survival probability at theend of 24 months was 0.54 (95% CI: 0.40 0.67).

Variables that were related to outcome are summarised inTable 1 for patients with tuberculosis-free survival, forpatients with treatment failure, relapse or death and forthose lost to follow-up during the study period. Thedistribution of cytokine responses to PPD stimulation wereskewed, with low or absent responses in the plasma of manypatients, especially IL-5 ( Table 2 ). The response to PPD wascorrelated with the response to mitogen for all three

cytokines (Spearmans rho 0.4989, p 0.0000; 0.4620, p 0.0000; and 0.7096, p 0.0000 for the IFN- g , IL-5 andIL-10 responses, respectively). The IFN- g response to PPDwas correlated with CD4+ lymphocyte count (Spearmansrho 0.2587, p 0.0059), the CD3+ lymphocyte count(Spearmans rho 0.2535, p 0.0070) and the IL-5 re-sponse to PPD (Spearmans rho 0.3514, p 0.0002), butnot the CD8+ lymphocyte count or the IL-10 response to PPD.Similarly, the IL-5 response to PPD correlated with the CD4+lymphocyte count (Spearmans rho 0.3378, p 0.0004)and the CD3+ lymphocyte count (Spearmans rho 0.2291, p 0.0182), but not the CD8+ count. The IL-10 response didnot correlate with any T-cell subset.

In multivariate analysis, the strongest predictor of a goodoutcome was the whole blood IFN- g response to PPDstimulation at 2 months ( p 0.002; Table 3 ; Figure 2). Thehazard ratio of 0.9996 reects the risk reduction associatedwith a 1pg/ml elevation in the IFN- g response; thus therelative hazard reduction for treatment failure, death orrelapse associated with a 1000 pg/ml elevation is 31%.Higher creatinine levels were also associated with morefavourable outcome ( p 0.001; Figure 3). A creatinineconcentration elevation of 5 mmol/l was associated with a26% reduction in the risk of treatment failure, death orrelapse by 24 months. Male sex was associated withunfavourable outcome ( Figure 4), although in contrast tothe associations of the IFN- g response and serum creatinine,

this nding was not supported by the univariate analysis.The associations of IFN- g response to PPD stimulation andcreatinine were preserved if the data were reanalysed usingthe assumption that the 62 patients lost to follow-up had agood clinical outcome at 24 months. Antitubercular drugconcentrations were not associated with the 24-monthoutcome. There was, however, correlation between rifam-picin exposure and the IFN- g response to PPD stimulation(Spearmans rho 0.235, p 0.013; Figure 5). A univariateanalysis including only those patients with the smallest andbiggest IFN-g responses to PPD, demonstrated that thosewith IFN-g responses in the lowest quartile ( p 447.5 pg/ml)had a risk of treatment failure, relapse or death 12.03 (95%CI: 1.5692.82) times greater than those with IFN- gresponses in the highest quartile ( 4 4554pg/ml). Rifampinlevels and IFN-g responses were positively correlated ( Figure 4).

ARTICLE IN PRESS

Rifampin levels, IFN- g release and outcome in complicated pulmonary tuberculosis 559


5/9


The median AUC0 8 for rifampin was signicantly reducedamongst those with IFN- g responses in the lowest quartilein comparison to the remaining patients (13.51mg h/l,95% CI: 6.2719.99 vs. 20.85 mg h/l, 95% CI: 17.64 22.74; p 0.0452).

Discussion

We have investigated factors determined at 2 months thatassociate with subsequent 24-month outcome amongsthospitalised patients with PTB in South Africa. The strongest

ARTICLE IN PRESS

Table 2 Summary of covariates in the study participants and the univariate associations of each with tuberculosis treatmentfailure, relapse or death during the study period using Cox proportional hazards models.

Covariate n Summary statistics Hazard ratio (95% condenceinterval) for univariate cox

proportional hazards models

p-value

IFN-g response to PPD (pg/ml) 112 1092 (448, 4554) 0.9997 (0.9994 0.9999) 0.002IL-10 response to PPD (pg/ml) 112 242 (68, 421) 0.9983 (0.9966 1.0000) 0.055IL-5 response to PPD (pg/ml) 106 0 (0, 17) 0.9994 (0.9951 1.0038) 0.795CD4+ lymphocyte count( 109 l 1)

112 0.690 (0.519, 0.924) 0.9996 (0.9985 1.0007) 0.503

CD8+ lymphocyte count( 109 l 1)

112 0.409 (0.299, 0.688) 0.9998 (0.9984 1.0011) 0.731

CD3+ lymphocyte count( 109 l 1)

112 1.298 (1.040, 1.663) 0.9997 (0.9989 1.0004) 0.394

Male sex 112 50 males (45%) 1.3524(0.63502.8805) 0.434Retreatment 112 71 (63%) 1.4647 (0.6621 3.2404) 0.346HIV infection 112 14 (12.5%) 2.004 (0.85054.7225) 0.112Age (years) 112 36 (29, 46) 0.9850 (0.9549 1.0160) 0.340Albumin (g/l) 112 34.5 (31, 38) 0.8860 (0.8257 0.9506) 0.001Creatinine (mol/l) 110 74 (65, 83) 0.9644 (0.9371 0.9926) 0.014Rifampicin AUC (mg h/l) 111 19.04 (9.80, 28.58) 1.0042 (0.9801 1.0289) 0.736Isoniazid AUC (mg h/l) 111 26.49 (20.38, 35.79) 1.0053 (0.9640 1.0484) 0.805Pyrazinamide AUC (mgh/l) 111 292.56 (245.64, 335.27) 1.0015 (0.9961 1.0070) 0.580Ethambutol AUC (mgh/l) 100 18.84 (15.87, 23.40) 1.009732 (0.9410 1.0835) 0.788

Expressed as median (interquartile range) unless indicated otherwise.

Table 1 The patient characteristics for those with tuberculosis-free survival at 24 months (good outcome), for patients withtreatment failure, relapse or death at 24 months (poor outcome) and for those lost to follow-up during the study period,respectively.

Covariate Poor outcome ( n 28) Good outcome ( n 22) Lost to follow-up ( n 62)

IFN-g response to PPD (pg/ml) 488 (168, 1729) 2642 (493, 4736) 1828 (584, 5505)IL-10 response to PPD (pg/ml) 131 (68, 236) 275 (127, 518) 299 (47, 456)IL-5 response to PPD (pg/ml) 0 (0, 11) 0 (0, 4) 0 (0, 20)CD4+ lymphocyte count ( 109 l 1) 624 (352, 960) 668 (522, 906) 706 (587, 909)CD8+ lymphocyte count ( 109 l 1) 387 (216, 627) 433 (305, 786) 437 (332, 690)CD3+ lymphocyte count ( 109 l 1) 1227 (803, 1534) 1276 (1046, 1742) 1369 (1121, 1723)Male sex 12 (43%) 7 (32%) 31 (50%)Retreatmentm y 19 (68%) 13 (59%) 39 (63%)HIV infectionz 7 (25%) 3 (14%) 4 (6%)Age (years) 34.5 (28.5, 41.5) 37 (26, 47) 37 (29,49)Albumin (g/l) 32 (26, 35) 35 (33, 38) 36 (33, 38)Creatinine ( mmol/l) 70 (61, 76) 81 (73, 90) 74 (65, 82)Rifampicin AUC (mg h/l) 15.3 (7.35, 28.15) 11.1 (4.6, 22.9) 20.8 (14.0, 31.7)

Isoniazid AUC (mg h/l) 28.3 (18.21, 36.85) 25.9 (22.6, 31.1) 24.9 (20.7, 35.8)Pyrazinamide AUC (mgh/l) 293.4 (238.1, 379.6) 280.1 (253.7, 310.2) 298.3 (247.1, 335.3)Ethambutol AUC (mg h/l) 19.8 (15.2, 22.6) 18.2 (15.8, 21.3) 19.8 (16.3, 24.3)

Expressed as median (interquartile range) unless indicated otherwise.yPreviously treated for tuberculosis or received 4 1 month of antitubercular treatment prior to admission.z All patients underwent voluntary counselling and testing. None of the patients were treated with antiretroviral drugs.



6/9


association was between the IFN- g response and PPD inwhole blood. Although drug levels were not associated withoutcome, the AUC 0 8 for rifampicin correlated with the IFN-g response, whereas the levels of isoniazid, pyrazinamideand ethambutol were not associated with this variable.

The form of IFN-g release assay that we employed differsin important respects from the commercially availableT-Spot. TB and QuantiFERON-TB assays that are beingextensively researched. Firstly, the stimulus we employedwas PPD which is not specic for M. tuberculosis andobviously contains a wider range of antigens than the regionof difference 1 antigens (ESAT-6 and CFP-10) upon which thecommercial assays rely. Secondly, commercial detection of

ARTICLE IN PRESS

0 5 10 15 20 25

analysis time

Kaplan-Meier survival estimates, by IFN_gamma

1.00

0.75

0.50

0.25

0.00

IFN_gamma = < 25th centile

IFN_gamma = 25th to < 50th centile

IFN_gamma = 50th to < 75th centile

IFN_gamma = />75th centile

Figure 2 KaplanMeier plot describing the proportion of patients remaining free of tuberculosis treatment failure, disease relapseor death during the study period, by the interferon- g response (interferon- g response o 25th centile [448 pg/ml]; interferon- gresponse X 25th centile, o 50th centile [448, o 1092 pg/ml]; interferon- g response X 50th centile, o 75th centile [1092, o 4554pg/ml]; and interferon- g response X 75th centile [4554 pg/ml]) after 2 months of inpatient treatment with rifampin based regimens.

0 5 10 15 20 25

1.00

0.75

0.50

0.25

0.00

Kaplan-Meier survival estimates, by creatinine

analysis time

creatinine = < 25th centile

creatinine = 25th to < 50th centilecreatinine = 50th to < 75th centile

creatinine = />75th centile

Figure 3 KaplanMeier plot describing the proportion of patients remaining free of tuberculosis treatment failure, disease relapseor death during the study period, by the serum creatinine concentration (serum creatinine below 25th centile [65 mmol/l]; serumcreatinine X 25th centile, o 50th centile [65, o 74 mmol/l]; serum creatinine X 50th centile, o 75th centile [74, o 83 mmol/l]; andserum creatinine X 75th centile [83 mmol/l]) after 2 months of inpatient treatment with rifampin based regimens.

Table 3 The multivariate Cox regression model bestdescribing the proportional hazards of covariate factorsassociated (at the signicance level of 0.05) withtuberculosis treatment failure, relapse or death duringthe study period.

Covariate factor Hazard ratio (95%condence interval)

p-value

IFN-g response to PPD(pg/ml)

0.9996 (0.9994, 0.9999) 0.002

Creatinine ( mmol/l) 0.9406 (0.9081, 0.9742) 0.001Male sex 3.0983 (1.2466, 7.7007) 0.015



7/9


8/9


component of the PPD stimulated response arises frominnate cells.

An interesting association between serum creatinineand positive outcome was observed. Most participantshad low creatinine levels (90% of participants had creatininein the range 52 97 mmol/l; normal range: 75 115 mmol/l)and few patients had chronic renal impairment which onewould expect to be associated with a poor rather thangood outcome. We therefore ascribe the relationship to thehigher creatinine values reecting better nutritionalstatus, in particular protein balance. Protein energy mal-nutrition predisposes to experimental tuberculosis 20 andtuberculosis is a well-recognised associate of wastingsyndromes in humans. It is difcult therefore to quantita-tively dissect cause from effect in human studies but ourdata support the idea that protein supplementation duringthe course of therapy of tuberculosis may improve theoutcome.

Patients able to mount a stronger IFN- g response after 2

months of treatment also had higher AUC 0 8 values forrifampin. Whether greater bioavailability of rifampin pro-motes a stronger IFN- g response is not clear from this study.The study was underpowered to detect any effect of rifampin concentration on outcome at 24 months. Moreover,the immune response may be more important than thesterilising activity of the antitubercular regimen for pre-vention of disease recurrence in the months followingtreatment in a community with high rates of re-infection.

Our study has a number of limitations. Firstly, the rates of relapse and death might be exaggerated due to thesubstantial number of censored individuals (62 or 55%) whomay have had a good treatment outcome. Notwithstandingthis, 25% of the 112 patients are known to have failedtreatment, relapsed or died within the 24-month periodfollowing their admission to the hospital. The overalloutcome for the region (Boland/Overberg) of patients at24 months after treatment initiation is not known, but theproportion of retreatment patients in our study was high,being approximately equal to the number of patientsreceiving treatment for the rst time ( Table 2 ). Factorstherefore contributing to the poor outcome we observedundoubtedly included selection of patients intrinsically athigh risk of treatment failure (i.e. those who were admittedto Brewelskloof Hospital for a variety of reasons which mightpredispose to a poor prognosis, such as poor treatmentadherence, severe or complicated disease, concomitant

illness and malnutrition, failure to respond to clinic-basedtreatment). The region has a very high incidence of tuberculosis ( 4 1%) and it is also possible that re-infectionwith a new strain of M. tuberculosis could have occurred in asubstantial proportion of patients as has been shown inanother Western Cape population with a lower overallincidence of tuberculosis. 21 The radiographic extent of disease was not measured. The extents to which IFN- gresponse and serum creatinine were associated with out-come independently of disease extent were therefore notaddressed. The analysis may also have been underpoweredfor some covariates: for example highly variable continuouscovariates (e.g. rifampicin levels), or dichotomous covari-ates with only a few subjects in one group (e.g. HIV-infection status). As drug concentrations were measuredduring a single dosing interval, variability attributable to

adherence was not accounted for. Therefore, the resultsshould not be generalised without caution.

However, our study supports and extends observationsthat suggest that an immune component may contribute tosuccessful resolution of both latent and active tuberculosis.An important role for IFN- g , a cytokine that is recognised tobe crucial to defence against mycobacteria, appears tounderpin this. The possibility that the cytokine could beused as a surrogate marker to select persons at risk of a pooroutcome for further intensive intervention is also raised anddeserves investigation.

Acknowledgements

We thank the staff and patients of Brewelskloof Hospital;Professors Gary Maartens and Peter Smith of the Division of Clinical Pharmacology, University of Cape Town for assis-tance with this work and helpful discussions; Dr. MustaphaKeraan of the Division of Clinical Immunology, Departmentof Medicine, Groote Schuur Hospital and University of CapeTown, for performing ow cytometric analysis.

Funding: The Medical Research Council of South Africasupported this project through a Self-initiated ResearchGrant. RJW holds a Wellcome Trust Senior Fellowship(072070).

Competing interests: None of the authors are aware of anypotential conict of interest.

Ethical approval: The study was approved by the Universityof Cape Town Research Ethics Committee (REC 122/99), wasconducted in accordance with the Helsinki Declaration of 1975 as revised in 1983 and was performed in collaborationwith the regional health authorities.

References

1. Zignol M, Hosseini MS, Wright A, et al. Global incidence of multidrug-resistant tuberculosis. J Infect Dis 2006;194 :47985.

2. Chang KC, Leung CC, Yew WW, Ho SC, Tam CM. A nestedcase control study on treatment-related risk factors for earlyrelapse of tuberculosis. Am J Respir Crit Care Med 2004;170 :112430.

3. Leung CC, Li T, Lam TH, et al. Smoking and tuberculosis amongthe elderly in Hong Kong. Am J Respir Crit Care Med 2004;170 :102733.

4. Coker R, McKee M, Atun R, et al. Risk factors for pulmonarytuberculosis in Russia: case control study. BMJ 2006;332 :857.

5. Falzon D, Le Strat Y, Belghiti F, Infuso A, EuroTB Correspon-dents. Exploring the determinants of treatment success fortuberculosis cases in Europe. Int J Tuberc Lung Dis 2005;9 :12249.

6. Singhal M, Banavalikar JN, Sharma S, Saha K. Peripheral blood Tlymphocyte subpopulations in patients with tuberculosis andthe effect of chemotherapy. Tubercle 1989;70 :1718.

7. Chan CH, Lai CK, Leung JC, Ho AS, Lai KN. Elevated interleukin-2 receptor level in patients with active pulmonary tuberculosisand the changes following anti-tuberculosis chemotherapy. Eur Respir J 1995;8 :703.

8. Wilkinson RJ, Vordermeier HM, Wilkinson KA, et al. Peptide-specic T cell response to Mycobacterium tuberculosis : clinical

ARTICLE IN PRESS



9/9


spectrum, compartmentalization, and effect of chemotherapy. J Infect Dis 1998;178 :7608.

9. Lalvani A, Pathan AA, McShane H, et al. Rapid detection of Mycobacterium tuberculosis infection by enumeration of anti-gen-specic T cells. Am J Respir Crit Care Med 2001;163 :8248.

10. De Jong BC, Israelski DM, Corbett EL, Small PM. Clinicalmanagement of tuberculosis in the context of HIV infection. Annu Rev Med 2004;55 :283301.

11. Wilkinson KA, Kon OM, Newton SM, et al. Effect of treatment of latent tuberculosis infection on the T cell response toMycobacterium tuberculosis antigens. J Infect Dis 2006;193 :3549.

12. Ewer K, Millington KA, Deeks JJ, Alvarez L, Bryant G, Lalvani A.Dynamic antigen-specic T cell responses after point-sourceexposure to Mycobacterium tuberculosis . Am J Respir Crit CareMed 2006;174 :8319.

13. Carrara S, Vincenti D, Petrosillo N, Amicosante M, Girardi E,Goletti D. Use of a T cell-based assay for monitoring efcacy of antituberculosis therapy. Clin Infect Dis 2004;38 :7546.

14. McIlleron H, Wash P, Burger A, Norman J, Folb PI, Smith P.Determinants of rifampicin, isoniazid, pyrazinamide andethambutol pharmacokinetics in a cohort of tuberculosispatients. Antimicrob Agents Chemother 2006;50 :11707.

15. Lalvani A, Nagvenkar P, Udwadia Z, et al. Enumeration of Tcellsspecic for RD1-encoded antigens suggests a high prevalence of latent Mycobacterium tuberculosis infection in healthy urbanIndians. J Infect Dis 2001;183 :46977.

16. Hirsch CS, Toossi Z, Othieno C, et al. Depressed T-cellinterferon-gamma responses in pulmonary tuberculosis: analysisof underlying mechanisms and modulation with therapy. J InfectDis 1999;180 :206973.

17. Seah GT, Scott GM, Rook GA. Type 2 cytokine gene activationand its relationship to extent of disease in patients withtuberculosis. J Infect Dis 2000;181 :3859.

18. Boussiotis VA, Tsai EY, Yunis EJ, et al. IL-10-producing T cellssuppress immune responses in anergic tuberculosis patients. J Clin Invest 2000;105 :131725.

19. Dieli F, Friscia G, Di Sano C, et al. Sequestration of Tlymphocytes to body uids in tuberculosis: reversal of anergyfollowing chemotherapy. J Infect Dis 1999;180 :2258.

20. Chan J, Tian Y, Tanaka KE, et al. Effects of protein caloriemalnutrition on tuberculosis in mice. Proc Natl Acad Sci USA1996;93 :1485761.

21. Verver S, Warren RM, Beyers N, et al. Rate of reinfectiontuberculosis after successful treatment is higher than rate of new tuberculosis. Am J Respir Crit Care Med 2005;171 :14305.

ARTICLE IN PRESS


mcilleron tuberculosis 2007

Documents