J Antimicrob Chemother 2011; 66: 1687 1695

doi:10.1093/jac/dkr210 Advance Access publication 3 June 2011

Pulmonary resection for patients with multidrug-resistant

tuberculosis: systematic review and meta-analysis
Hong-Bin Xu 1*, Rui-Hua Jiang 2 and Ling Li 1
1

Department of Clinical Pharmacy, Shanghai Tenth Peoples Hospital, Tongji University School of Medicine, Shanghai 200072, China;
2
Department of Tuberculosis, Shanghai Pulmonary Hospital, Tongji University School of Medicine, Shanghai 200433, China
*Corresponding author. Tel: +86-2166302761; Fax: +86-2166307668; E-mail: xuhongbin119@yahoo.cn
These authors contributed equally to this work.

Received 2 March 2011; returned 7 April 2011; revised 27 April 2011; accepted 1 May 2011
Downloaded from http://jac.oxfordjournals.org/ by guest on April 20, 2014

Objectives: Multidrug-resistant tuberculosis (MDR-TB) has become an emerging global public health crisis.
Several studies have suggested that pulmonary resection has efficacy in the treatment of MDR-TB. A systematic
review of the available therapeutic studies was conducted to determine the treatment outcome among
patients with MDR-TB who underwent pulmonary resection.
Methods: To evaluate pulmonary resection for MDR-TB, a random-effect meta-analysis of the available studies
was used to assess the overall treatment outcome. Subgroup analyses were also conducted by separating
studies based on each characteristic independently.
Results: After screening 4996 articles, 15 clinical reports with a mean of 63 patients per report met the
inclusion criteria. Analysis of the studies showed that the estimated pooled treatment success rate of pulmonary resection for patients with MDR-TB was 84% [95% confidence interval (CI) 78% 89%]. The rates of failure,
relapse, death and default were 6% (95% CI 4%8%), 3% (95% CI 1%4%), 5% (95% CI 2% 8%) and 3%
(95% CI 1% 5%), respectively. The proportion of patients treated successfully did not differ significantly on
the basis of any of the individual study characteristics.
Conclusions: Substantial heterogeneity in the study characteristics prevented a more conclusive determination
of what factors had the greatest effect on the proportion of patients that achieve treatment success and
limited the validity of this analysis. Some important variables, including patient HIV status, were inconsistently
reported between studies. These results underscore the importance of strong patient support and treatment
follow-up systems to develop successful MDR-TB treatment programmes.
Keywords: pulmonary surgical procedures, MDR-TB, treatment outcome

Introduction
Tuberculosis (TB) is a major global health problem, with 9 million
new cases and almost 2 million deaths per year.1 Most patients
with TB can be successfully treated using short-course medical
chemotherapy, which consists of a four-drug regimen including
isoniazid, rifampicin, pyrazinamide and ethambutol.2 However,
a small proportion of patients with pulmonary TB require surgical
treatment.3 The indications for surgery usually include the management of complications of TB (including haemoptysis, bronchiectasis, bronchial stenosis, bronchopleural fistula and
aspergilloma) and the management of drug-resistant forms of
the disease.4 7 A wide variety of procedures have been reported,
including surgical resection and thoracoplasty.8 Patients usually
do well with surgery, with cure rates of 60% 100% being
achieved.9,10

Multidrug-resistant TB (MDR-TB), a major indication for

surgery, is defined as having in vitro resistance to at least both
isoniazid and rifampicin, the two most powerful existing
anti-TB agents, and has become an emerging global public
health crisis.11,12 Strains of MDR Mycobacterium tuberculosis are
often resistant to other anti-TB agents in addition to isoniazid
and rifampicin. When such an isolate is resistant to any secondline injectable agent (e.g. amikacin, capreomycin or kanamycin)
and any fluoroquinolone, the strain is termed extensively
drug-resistant TB (XDR-TB).13 In many countries the rates of
drug-resistant TB are increasing and more international efforts
have been mobilized to confront this emerging infectious
disease. An estimated 489 000 cases of MDR-TB occurred worldwide in 2006.14,15 The management of drug-resistant TB requires
extended treatment and expensive and potentially toxic drug
regimens, and often results in higher rates of treatment failure

# The Author 2011. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved.
For Permissions, please e-mail: journals.permissions@oup.com

1687

Systematic review

Methods
The meta-analysis was conducted in accordance with QUOROM
guidelines.32

Data sources
PubMed, Embase and the Cochrane Library for were searched for studies
through July 2010. There was no limit put on the earliest date of publication. The search terms included were as follows: tuberculosis; pulmonary resection; pneumonectomy; surgery; surgical treatment;
lung resection; drug-resistant; multidrug resistant; multi-drug resistant; MDR; and multiple drug-resistant. In addition, the online archives
of the International Journal of Tuberculosis and Lung Disease were
reviewed for applicable studies not found in the previous search. The
reference lists of identified original articles and reviews were searched
for other relevant articles. Abstracts, chapters of books, conference proceedings or correspondence were not included. The initial search had
no language restrictions, but studies not published in the English
language were excluded from the data extraction process. As very
small studies may be vulnerable to selection bias, only studies including
10 patients were selected for inclusion in the analysis.

Study selection
Two reviewers (H.-B. X. and R.-H. J.) selected articles in the following two
stages: titles and abstracts; and then full-text articles. Discrepancies
between the two reviewers were resolved by consensus or through discussion with a third reviewer (L. L.). Two reviewers (H. B. X. and R. H. J.)
performed data extraction for all the articles and a third reviewer (L. L.)
independently performed data extraction for one-third of the articles to
assess accuracy in the data extraction. Studies were required to meet
the following inclusion criteria: (i) confirmation that patients had
MDR-TB using drug-susceptibility testing of cultured M. tuberculosis; (ii)
treatment outcome definitions specified by mycobacterial culture endpoints; and (iii) outcomes reported according to WHO classifications of
success (cure or treatment completion), failure, relapse, death and
default (treatment interruption).30 Because this study was a systematic
review, ethics committee approval or written informed consent from
the participants was not required. The methodological quality of the
eligible studies was assessed mainly according to Cochrane-based
criteria.

1688

Data extraction
We recorded treatment outcomes according to WHO classifications of
treatment success (cure or treatment completion), failure, relapse,
death and default (treatment interruption).30 Patients still on treatment
who were classified as probable cure or probable failure were added to
the success and failure categories, respectively. Patients who remained
on treatment but were not assigned an interim outcome were not
included in this analysis. All patients who were classified as transfer
out were added to the default category.
For each study, data were gathered on patients characteristics, treatment protocols and study definitions. Patients characteristics included
previous TB treatment history, HIV prevalence, and the mean number
of drugs to which patients isolates were resistant. Data on treatment
protocols included the indication for surgery, pre-operative lung lesion
type and type of pulmonary resection. In addition, the length of
follow-up and definition of cure were recorded, if available. If no data
on the above-mentioned characteristics were reported in the primary
studies, the information was requested from the original authors.

Data synthesis and analysis

For each study, the yield of treatment outcomes including success,
failure, relapse, death and default was calculated. Pooled percentages
for each outcome from every study were included in a Bayesian
random-effects meta-analysis.33 A Bayesian model was chosen
because of its appealing nature, such as full allowance for all parameter
uncertainty in the model, the ability to include other pertinent information that would otherwise be excluded and the ability to extend the
model to accommodate more complex, but frequently occurring, scenarios. The Bayesian method overcomes some of the difficulties encountered with other traditionally used methods, such as wide variation in
sample sizes and estimates, the presence of covariates, and response
rates near 0 or 1. The Bayesian method directly interprets the credible
intervals of the posterior-effect estimates.34,35 The heterogeneity across
the studies was estimated by calculating I 2, an index of the proportion
of total variation across studies that is due to heterogeneity rather
than to chance.36 A P value of ,0.05 indicated that heterogeneity
among the group of studies being analysed was significant. The I 2 statistic was .50% (with P,0.05) for each treatment outcome (success,
failure, relapse, death and default). Therefore, a random-effects analysis,
incorporating the impact of both chance and heterogeneity among study
populations and study design, was chosen over the fixed-effects alternative, which assumes that differences among study outcomes are due
entirely to chance.
To examine the influence of each study characteristic individually on
the treatment success, subgroup analyses were also conducted by separating studies based on each characteristic independently. The individual
factors were ranked by the difference in treatment success between
reports that met the cut-off compared with all other reports. The differences in the proportion of patients achieving treatment success among
studies that met the cut-offs compared with those that did not were
then examined. Statistical analyses were done with WinBUGS and R. As
publication bias is a risk in meta-analyses, the potential presence of
this bias was tested for by using the Egger test.

Results
Study characteristics
As summarized in Figure 1, 15 studies (Table 1) were selected for
inclusion in the analysis,26 29,37 47 including 12 studies of pulmonary resection for patients with MDR-TB26 29,37 40,43 46 and
3 studies for patients with MDR-TB or XDR-TB.41,42,47 Because

Downloaded from http://jac.oxfordjournals.org/ by guest on April 20, 2014

and death compared with the drug-susceptible disease.16 18

Reported cure or treatment completion rates for MDR-TB using
second-line agents have ranged from 44% to 83%.19 24 Treatment failure is especially common in patients co-infected with
XDR-TB and HIV, where mortality approaches 100%.25
Medical treatment for drug-resistant TB with anti-TB drugs
does not achieve satisfactory outcomes. However, in several
studies, the resection of cavitary lesions or destroyed lobes
that harbour great numbers of the TB organism improved
patient outcome in cases with MDR-TB. Pulmonary resection
combined with antituberculous chemotherapy for MDR-TB has
shown success rates of 89% 96%.16,26 29 Recently, treatment
guidelines for MDR-TB were published. The role of surgery in
the treatment of MDR-TB continues to be controversial.30,31 We
conducted a systematic review and meta-analysis of the available therapeutic studies to determine the outcome among
patients with MDR-TB who were treated with pulmonary
resection.

JAC

Systematic review

2383 titles identified from literature search

2301 excluded based on

screening of title and abstract

66 retrieved for reading of the

entire text (including 8 added
from references and reviews)

15 articles included in this meta-analysis (949 patients)

Figure 1. Schematic representation of the study selection process.

of language problems, four studies published in Lithuanian and

Chinese were excluded. The reports came from several countries
and included treatment outcomes for a mean of 63 patients
per report (range 19172). Of the 15 studies, all
reported data on post-operative cure, 13 reported failure
rates,26 29,37 43,45,47 11 reported relapse rates,26,27,29,37 39,42 46 8
reported death rates26,28,29,37,41,43,44,47 and 5 reported default
studies
reported
data
on
rates.26,28,37,41,46 Thirteen
follow-up26,27,29,37 40,42 47 and only 4 (27%) reported ,10%
of patients lost to follow-up after surgery in the analysis.29,42,43,47 The length of follow-up after pulmonary resection
often varied and ranged from 3 to 204 months. Twelve studies
reported varied post-operative medical therapy regimens.26,27,29,37 40,42,44 47
Characteristics of the patient population and the cure definition differed among the included studies (Table 1). The
number of patients previously medically treated for TB ranged
from 95% to 100% in the 15 reviewed reports and in five
studies patients received a fluoroquinolone.41,44 47 Six (40%) of
the studies reviewed were published before 199526,28,37 39,43
and nine (60%) were published in 1995 or later.27,29,40 42,44 47
By the World Bank classification system, six (40%) studies were
from developing countries37,39,40,43 45 and nine (60%) were
from developed countries.26 29,38,41,42,46,47 Fourteen studies
reported post-operative complications.26,27,29,37 47 The preoperative lung lesion type was not consistent in the 11 studies
in which it was reported and was therefore inappropriate for
analysis.26,27,29,38,39,41,43 47 Major complications included
empyema in 9 studies, bleeding in 12, bronchopleural fistula in
8, air leakage in 8 and wound infection in 6. The results of methodological quality assessment of the studies are described in
Table 2. All studies are case series. Thirteen studies were

retrospective26,27,28,37 41,43 47 and 2 were prospective.29,42 The

statistical analysis methodology was well described in seven
studies,28,29,37,38,41,45,47 but was not reported in eight
studies.26,27,39,40,42 44,46

Treatment outcomes
Across all studies, the estimated pooled treatment success rate,
defined as the proportion of patients who were cured or completed treatment, was 84% [95% confidence interval (CI)
78% 89%]. The pooled treatment failure, relapse, death and
default rates were 6% (95% CI 4%8%), 3% (95% CI 1%
4%), 5% (95% CI 2%8%) and 3% (95% CI 1% 5%), respectively (Figure 2). However, the heterogeneity in the study characteristics led to significant variation in the reported treatment
outcome. Among the 12 reports of pulmonary resection for
patients with MDR-TB (mean 69 patients; range 23 172), the
mean proportion of patients achieving treatment success was
83% (95% CI 77% 89%; Figure 3 and Table 3). Among the
three studies involving patients with MDR-TB or XDR-TB (mean
42 patients; range 1972), the mean proportion achieving treatment success was higher than that for the studies with MDR-TB
patients, but the difference was not significant at 89% (95% CI
86% 92%; Figure 3 and Table 3).
We also analysed the effect of each of the other study characteristics independently (Table 3). The analysis showed that the
proportion of patients treated successfully did not differ significantly on the basis of any of the following individual study
characteristics: start year of the study; cure definition; national
income status; geographic region; or type of pulmonary resection
(Table 3). Differences in population characteristics, including

1689

Downloaded from http://jac.oxfordjournals.org/ by guest on April 20, 2014

51 excluded after review:

3 duplicate populations
5 case reports with fewer than 10 patients
6 had outcomes for MDR tuberculosis missing or incomplete
4 not original articles (letters, reviews, editorials)
4 not in English language
29 not a multidrug-resistant tuberculosis study

Study location

Years

Sample
size

Proportion
previously
treated (%)

HIV
prevalence

Median resistance
[n drugs (range)]

Pulmonary resection (type, n)

Definition of cure

Pomerantz
et al. 26
Park et al. 27

USA

1983 2000

172

100

NR

NR

pneumonectomy 82, lobectomy 98

South
Korea

1995 1999

49

100

4.7

Chan et al. 28

USA

1984 1998

130

100

NR

Somocurcio
et al. 29

USA

1999 2004

121

95

1%

7 (3 11)

Leuven
et al. 37

South
Africa

1990 1995

61

100

Sung et al. 38

South
Korea

1994 1998

27

100

NR

4.4 (2 10)

Chiang
et al. 39

China

1990 1999

26

100

NR

3.9 (3 6)

Naidoo
et al. 40

South
Africa

1996 2000

23

100

17%

NR

Kwon et al. 41 South

Korea
Park et al. 42
South
Korea
Kir et al. 43
Turkey

1995 2004

35

100

NR

5 (3 6)

1998 2004

19

100

4.7 (2 8)

pneumonectomy 12, lobectomy 35, 3 months of

wedge resection 1, cavernoplasty
consecutive
1
negative cultures
pneumonectomy 66, lobectomy 64 3 months of
consecutive
negative cultures
pneumonectomy 27, lobectomy 87, 6 months of
segmentectomy or wedge
consecutive
resection 3, other 4
negative cultures
pneumonectomy 35, lobectomy 26, 3 months of
segmentectomy 1
consecutive
negative cultures
pneumonectomy 9, lobectomy 16, 3 months of
segmentectomy 2
consecutive
negative cultures
pneumonectomy 10, lobectomy 13, culture negative
segmentectomy and/or wedge
resection 4
pneumonectomy 17, lobectomy 6
18 months of
consecutive
negative cultures
pneumonectomy 14, lobectomy 20, Laserson definition
segmentectomy 1
pneumonectomy 1, lobectomy 18
culture negative

1993 2005

79

100

NR

Mohsen
Egypt
et al. 44
Wang et al. 45 China

1995 2005

23

100

1995 2006

56

100

Japan

2000 2007

56

South
Korea

1996 2008

72

Shiraishi
et al. 46
Kang et al. 47

culture negative

Follow-up
(months)
4 204
6 36

NR

19 103

24

3 45

15 80

18

NR
68
18 24

4 (3 6)

pneumonectomy 43, lobectomy 37, culture negative for

segmentectomy 1
18 consecutive
months
pneumonectomy 11, lobectomy 12 culture negative

3.5 (2 4)

pneumonectomy 25, lobectomy 31

9 137

100

5.6 (2 10)

100

6 (5 8)

2 months of
consecutive
negative cultures
pneumonectomy 22, lobectomy 33, culture negative
segmentectomy 6
pneumonectomy 23, lobectomy 38, Laserson definition
segmentectomy 10, wedge
resection 1

NR, not reported.

14 27

8 105
18 24

Systematic review

Downloaded from http://jac.oxfordjournals.org/ by guest on April 20, 2014

1690

Table 1. Summary of included studies

JAC

Systematic review

prevalence of HIV, and mean number of resistant drugs also did

not lead to significantly different outcomes.

Table 2. Assessment of study quality

Study
location

Incomplete
outcome data
addressed

Prospective

USA

NR

yes

no

South Korea
USA
USA

NR
NR
NR

yes
no
yes

no
no
yes

South Africa

NR

yes

no

South Korea
China

consecutive
NR

yes
yes

no
no

South Africa

NR

yes

no

South Korea
South Korea
Turkey
Egypt

NR
NR
NR
consecutive

no
yes
no
yes

no
yes
no
no

China
Japan

consecutive
consecutive

yes
yes

no
no

South Korea

NR

yes

no

NR, not reported.

Publication bias
The result of the Egger test for publication bias was significant
(P 0.020). The funnel plots for publication bias (Figure 4) also
showed asymmetry. These results indicate a presence of
potential for publication bias.

Discussion
MDR-TB ultimately develops from the inadequate treatment of
active pulmonary TB. There are multiple reasons for inadequate
therapy; poor prescribing practices with insufficient treatment
duration and poor drug selection are well-recognized contributors.48,49 Systemic problems, through inadequate public health
resources and unpredictable drug supplies, also play a role.50
In addition, irregular medication intakewhether due to insufficient patient education, adverse events or socioeconomic determinantscontributes to resistance. Also, a significant proportion
of patients acquire the drug-resistant disease because they live
in an environment with a high prevalence of drug-resistant TB.
The medical treatment of MDR-TB is complex and costly, and
is associated with relatively low cure rates, high relapse,
increased toxicity, and high morbidity and mortality. Two previous systematic reviews evaluated the existing evidence regarding medical treatment regimens for MDR-TB. The authors used a
meta-analysis of the available therapeutic studies to assess
treatment outcomes. The results of the two reviews showed
that the pooled treatment success estimate, defined as the
Success (95% CI)

al.26

Pomerantz et
Park et al.27
Chan et al.28
Somocurcio et al.29
Leuven et al.37
Sung et al.38
Chiang et al.39
Naidoo et al.40
Kwon et al.41
Park et al.42
Kir et al.43
Mohsen et al.44
Wang et al.45
Shiraishi et al.46
Kang et al.47

92 (8796)
94 (8399)
76 (6883)
63 (5471)
77 (6587)
74 (5489)
88 (7098)
96 (78100)
89 (7397)
84 (6097)
89 (7995)
91 (7299)
88 (7695)
91 (8097)
90 (8196)

Summary

84 (7889)
I 2 = 79

20

60
80
40
Treatment success (%)

100 (P < 0.001)

Failure (%)

Relapse (%) Death (%)

Default (%)

2
4
7
11
10
19
8
4
9
5
1
0
9
0
6

2
2
0
1
10
7
4
0
0
5
1
4
4
9
0

6 (48)

3 (14)

5 (28)
I 2 = 77

I 2 = 73

(P = 0.004)

(P = 0.003)

(P < 0.001)

(P < 0.001)

I 2 = 56

I 2 = 57

3
0
8
18
3
0
0
0
3
0
3
4
0
0
1

3
0
6
0
5
0
0
0
7
0
0
0
0
9
0

3 (15)

Figure 2. Treatment success and other treatment outcomes for pulmonary resection of multidrug-resistant tuberculosis. Treatment effects and
summaries were calculated using a random-effects model weighted by study population. The 95% confidence intervals (95% CI) are shown.

1691

Downloaded from http://jac.oxfordjournals.org/ by guest on April 20, 2014

Pomerantz
et al. 26
Park et al. 27
Chan et al. 28
Somocurcio
et al. 29
Leuven
et al. 37
Sung et al. 38
Chiang
et al. 39
Naidoo
et al. 40
Kwon et al. 41
Park et al. 42
Kir et al. 43
Mohsen
et al. 44
Wang et al. 45
Shiraishi
et al. 46
Kang et al. 47

Sampling
method

Systematic review

Success (95% CI)

(a)
Pomerantz et al.26
Park et al.27
Chan et al.28
Somocurcio et al.29
Leuven et al.37
Sung et al.38
Chiang et al.39
Naidoo et al.40
Kir et al.43
Mohsen et al.44
Wang et al.45
Shiraishi et al.46

92 (8796)
94 (8399)
76 (6883)
63 (5471)
77 (6587)
74 (5489)
88 (7098)
96 (78100)
89 (7995)
91 (7299)
88 (7695)
91 (8097)

83 (7789)
I 2 = 83
0

20

40

60

80

100

P < 0.001

Treatment success (%)

Success (95% CI)
(b)
Kwon et al.41
Park et al.42
Kang et al.47

89 (7397)
84 (6097)
90 (8196)

Summary

89 (8692)
I2=0
(P < 0.768)

20

40

60

80

100

Treatment success (%)

Figure 3. Treatment success by type of study. (a) Treatment success for studies with multidrug-resistant tuberculosis patients. (b) Treatment success
for studies with multidrug-resistant and extensively drug-resistant tuberculosis patients. Treatment effects and summaries were calculated using a
random-effects model weighted by study population. The 95% confidence intervals (95% CI) are shown.

proportion of patients who were cured or completed treatment,

was only 62%, while the pooled default, death, and transferred
out rates were 13%, 11% and 2%, respectively.51,52
Because anti-TB medical therapy occasionally proves
inadequate to cure MDR-TB, adjunctive pulmonary resection
has been advocated in selected patients who have localized
disease, such as a cavitary lesion or a single destroyed lobe.53
The rationale for pulmonary resection is based on the belief
that pulmonary cavities harbour millions or billions of the
resistant organisms. The infection in these locations cannot be
controlled by the host-specific immunity and, consequently,
the surgical removal of these cavities may permit cure.54 Here,
a systematic review and meta-analysis is reported that evaluated the treatment outcomes of pulmonary resection for
patients with MDR-TB. The results showed that the pooled
success rate was 84%, and the pooled failure, relapse, death
and default rates were 6%, 3%, 5% and 3%, respectively.
Although pulmonary resection was an adjunct to medical treatment for patients with MDR-TB, it seemed that the overall

1692

treatment success rate of pulmonary resection was better than

that for medical treatment regimens. The overall incidence of
failure, relapse, death and default with pulmonary resection
was also relatively lower. To date, there has been no randomized
study to address the comparative efficacy of chemotherapy
versus surgery combined with chemotherapy in the management of MDR-TB. This type of study is urgently needed in the
future. Prospective randomized controlled trials comparing different treatment strategies for MDR-TB will provide greater insight
into the improvement of MDR-TB treatment protocols.
To determine which patient and programme characteristics
facilitate the greatest treatment success, data from the
reviewed 15 studies that included treatment outcomes for a
total of 949 patients receiving pulmonary resection were analysed. Among the studies that included MDR-TB and XDR-TB
patients receiving pulmonary resection, the treatment success
was relatively better than in studies including MDR-TB patients,
but the difference was not significant. In fact, no individual
patient or programme characteristic was associated with a

Downloaded from http://jac.oxfordjournals.org/ by guest on April 20, 2014

Summary

JAC

Systematic review

Table 3. Pooled treatment success among subgroups of studies

Heterogeneity
Treatment success (95% CI)

HIV prevalence
0%27,37,42 47
.0% and not reported26,28,29,38 41

8
7

88% (84 92)

81% (71 90)

0.250
,0.001

23
88

Average resistanceb
4.928,29,41,46,47
,4.927,37 39,42,44,45
not reported26,40,43

5
7
3

78% (67 89)

85% (80 91)
92% (89 95)

,0.001
0.126
0.453

87
40
0

Start year of study

1995 or later27,29,40 42,44 47
1994 or earlier26,28,37 39,43

9
6

83% (74 92)

84% (78 91)

,0.001
,0.001

83
77

5
10

80% (67 93)

86% (81 91)

,0.001
0.0012

89
67

National income status

developed countries26 29,38,41,42,46,47
developing countries37,39,40,43 45

9
6

83% (75 91)

87% (82 91)

,0.001
0.213

86
30

Geographic region
Asia27,37,39,41,42,45 47
Africa38,40,44
Europe and America26,28,29,43

8
3
4

89% (85 92)

84% (73 96)
80% (67 94)

0.440
0.046
,0.001

0
6
93

Type of pulmonary resectionc

pneumonectomy 50%28,37,40,43
pneumonectomy ,50%26,27,29,38,39,41,42,44 47
lobectomy 50%26,27,29,38,41,42,44 47
lobectomy ,50%28,37,39,40,43

4
11
10
5

81% (74 89)

85% (78 92)
85% (77 92)
82% (75 88)

0.017
,0.001
,0.001
0.025

71
82
84
64

Type of studiesd
studies including MDR patients26 29,37 40,43 46
studies including MDR +XDR patients41,42,47

12
3

83% (77 89)

89% (86 92)

,0.001
0.768

83
0.0

Definition of cure
5 months of negative cultures required29,40,41,43,47
,5 months of negative cultures required26 28,37 39,42,44 46

I 2 (%)

All 15 reports were included and classified by cohort factor.

Reports were classified on the basis of the mean number of drugs.
c
Reports were classified on the basis of the type of pulmonary resection in each study.
d
Reports were classified on the basis of studies involving patients with multidrug-resistant tuberculosis, or multidrug-resistant and extensively
drug-resistant tuberculosis.
b

significantly greater proportion of patients achieving treatment

success. However, substantial heterogeneity in the patient
populations, study designs and reporting limit the scope of
this analysis. It prevents a more conclusive determination of
what factors have the most effect on the proportion of patients
that achieve treatment success and limits the validity of this
analysis. The treatment protocols, and the reporting of key
patient and study characteristics were found to be inconsistent
across studies. Several studies omitted the pre-operative lung
lesion type and the average number of drugs to which each
patients TB isolate was resistant. HIV infection, which has
been associated with poor outcomes among patients receiving
treatment for MDR-TB and XDR-TB, was not assessed in five of
the studies reviewed.26,28,38,39,41 This absence, despite the confluence of the HIV and TB epidemics, exemplifies the

heterogeneity in TB epidemics and associated treatment strategies. Thus, the lack of significant findings associated with
certain variables in this analysis may be due to reporting insufficiency, rather than the absence of a real association. Furthermore, these variations in the recording of data necessary for
assessing treatment outcomes underscore the need for standardized data collection and reporting in programmes and
studies of MDR-TB, possibly through the use of an international
registry of treatment outcomes.56 The absence of a registry of
individual treatment outcomes or outcome studies precluded
the possibility of estimating publication bias in this
meta-analysis. Standards in outcome reporting are particularly
important in light of the emergence of XDR-TB.
Our systematic review and meta-analysis had several
strengths. The comprehensive search strategy enabled us to

1693

Downloaded from http://jac.oxfordjournals.org/ by guest on April 20, 2014

Reports (n)a

Systematic review

Acknowledgements

Log OR

We thank all patients who had been treated in these trials.

Funding
This work was supported by the Scientific Research Foundation of Shanghai Pharmaceutical Society (Grant No. 2009-YY-01-13).

2
0

0.5
SE of log OR

review articles from multiple databases though July 2010.

Moreover, two reviewers independently and reproducibly completed screening, study selection and data extraction. An
additional strength was the subgroup analysis to examine
the impact of each study characteristic individually on treatment outcome. This review also had several limitations. First,
the most important is our inability to identify any randomized
control trials. Instead, observational data were relied on exclusively for treatment outcomes. This likely introduced confounding to our pooled analysis, since crude outcomes, rather than
adjusted odds ratios, were reported for most trials. Second,
although our search strategy was systematic, we were not
able to include non-English-language papers, thereby limiting
the scope of the included studies. Analyses were limited by
the small number of included studies for a particular
outcome. Third, there was substantial variability in study
characteristics, which resulted in substantial heterogeneity,
even in stratified analyses. The pooled treatment outcome
estimates should be viewed with caution. Additionally, the preoperative lung lesion type was not reported precisely in most
studies; thereby, the influence of lesion type on treatment
outcome could not be analysed.
As in any systematic review, publication bias was a concern.
However, existing methods such as funnel plots and regression
asymmetry tests for publication bias were designed for the
meta-analysis of randomized controlled trials. To our knowledge,
they cannot be used to detect publication bias in a meta-analysis
of diagnostic studies or rates.57
The Stop TB strategy developed by WHO set the goal of
curing 85% of all detected TB cases by 2005.58 MDR-TB has presented challenges to achieving this objective in many areas.
Nonetheless, appropriate pulmonary resection might lead to
treatment success more readily and should be further investigated in current trials of MDR-TB therapy. In the absence of randomized clinical trials, more systematic documentation of
programmatic components and outcomes of MDR-TB treatment
can strengthen the evidence base for treatment. The comprehensive treatment of MDR-TB is of vital importance in promoting
public health to slow the spread and reduce the impact of
drug-resistant TB around the world.

1694

None to declare.

References
1 WHO. Global Tuberculosis Control 2009: Epidemiology, Strategy,
Financing. WHO/ HTM/TB/2009.411. http://www.who.int/entity/tb/
publications/global_report/2009/pdf/full_report.pdf (1 March 2011, date
last accessed).
2 American Thoracic Society, Centers for Disease Control and Prevention,
Infectious Diseases Society of America. Treatment of tuberculosis. Am J
Respir Crit Care Med 2003; 167: 60362.
3 Freixinet JG, Rivas JJ, Rodriguez de Castro F et al. Role of surgery in
pulmonary tuberculosis. Med Sci Monitor 2002; 8: CR782 6.
4 Souilamas R, Riquet M, Barthers F et al. Surgical treatment of active and
sequelar forms of pulmonary tuberculosis. Ann Thorac Surg 2001; 71:
4437.
5 Picciocchi A, Granone P, Margaritora S et al. Surgical management of
pulmonary tuberculosis. Rays 1998; 23: 193202.
6 Treasure R, Seaworth B. Current role of surgery in Mycobacterium
tuberculosis. Ann Thorac Surg 1995; 59: 1405 9.
7 Massard G, Dabagh A, Wihlm J et al. Pneumonectomy for chronic
infection is a high-risk procedure. Ann Thorac Surg 1996; 62: 1033 8.
8 Pomerantz M, Mault J. History of resectional surgery for tuberculosis
and other mycobacterial infections. Chest Surg Clin North Am 2000; 10:
1313.
9 Perelman M, Strelzov V. Surgery for pulmonary tuberculosis. World J
Surg 1997; 21: 45767.
10 Lahiri T, Agrawal D, Gupta R et al. Analysis of status of surgery in
thoracic tuberculosis. Indian J Chest Dis Allied Sci 1998; 40: 99 108.
11 Iseman MD. Treatment of multidrug-resistant tuberculosis. N Engl J
Med 1993; 329: 784 91.
12 Nachega JB, Chaisson RE. Tuberculosis drug resistance: a global
threat. Clin Infect Dis 2003; 36 (Suppl 11): S24 30.
13 Centers for Disease Control and Prevention. Revised definition of
extensively drug-resistant tuberculosis. MMWR 2006; 55: 1176.
14 Wright A, Zignol M. Anti-tuberculosis Drug Resistance in the World:
Fourth Global Report. Geneva: WHO, 2008.
15 Nathanson E, Gupta R, Huamani P et al. Adverse events in the
treatment of multidrug-resistant tuberculosis: results from the
DOTS-Plus initiative. Int J Tuberc Lung Dis 2004; 8: 1382 4.
16 Tahaoglu K, Torun T, Sevim T et al. The treatment of
multidrug-resistant tuberculosis in Turkey. N Engl J Med 2001; 19: 170 4.
17 Iseman MD. Management of multidrug-resistant tuberculosis.
Chemotherapy 1999; 45 (Suppl 2): 3 11.
18 Centers for Disease Control and Prevention. Extensively drug resistant
tuberculosisUnited States, 1993 2006. MMWR 2007; 56: 2503.

Downloaded from http://jac.oxfordjournals.org/ by guest on April 20, 2014

Figure 4. Funnel graph for the assessment of potential publication bias.

The funnel graph plots the log of the odds ratio (OR) against the standard
error (SE) of the log of the OR (an indicator of sample size). Circles
represent each study in the meta-analysis. The result of the Egger test
for publication bias was significant (P0.020).

Transparency declarations

Systematic review

19 Caminero JA. Treatment of multidrug-resistant tuberculosis: evidence

and controversies. Int J Tuberc Lung Dis 2006; 10: 82937.

JAC
39 Chiang CY, Yu MC, Bai KJ et al. Pulmonary resection in the treatment
of patients with pulmonary multidrug-resistant tuberculosis in Taiwan.
Int J Tuberc Lung Dis 2001; 5: 272 7.
40 Naidoo R, Reddi A. Lung resection for multidrug-resistant tuberculosis.
Asian Cardiovasc Thorac Ann 2005; 13: 172 4.

21 Nathanson E, Lambregts-van Weezenbeek C, Rich ML et al.

Multidrug-resistant tuberculosis management in resource-limited
settings. Emerg Infect Dis 2006; 12: 1389 97.

41 Kwon YS, Kim YH, Suh GY et al. Treatment outcomes for

HIV-uninfected patients with multidrug-resistant and extensively
drug-resistant tuberculosis. Clin Infect Dis 2008; 47: 496 502.

22 Suarez PG, Floyd K, Portocarrero J et al. Feasibility and

cost-effectiveness of standardised secondline drug treatment for
chronic tuberculosis patients: a national cohort study in Peru. Lancet
2002; 359: 1980 9.

42 Park SK, Kim JH, Kang H et al. Pulmonary resection combined with
isoniazid- and rifampin-based drug therapy for patients with
multidrug-resistant and extensively drug-resistant tuberculosis. Int J
Infect Dis 2009; 13: 1705.

23 Mitnick C, Bayona J, Palacios E et al. Community-based therapy for

multidrug-resistant tuberculosis in Lima, Peru. N Engl J Med 2003; 348:
11928.

43 Kir A, Inci I, Torun T et al. Adjuvant resectional surgery improves cure

rates in multidrug-resistant tuberculosis. J Thorac Cardiovasc Surg 2006;
131: 6936.

24 Palmero DJ, Ambroggi M, Brea A et al. Treatment and follow-up of

HIV-negative multidrug-resistant tuberculosis patients in an infectious
diseases reference hospital, Buenos Aires, Argentina. Int J Tuberc Lung
Dis 2004; 8: 77884.

44 Mohsen T, Zeid AA, Haj-Yahia S. Lobectomy or pneumonectomy for

multidrug-resistant pulmonary tuberculosis can be performed with
acceptable morbidity and mortality: a seven-year review of a single
institutions experience. J Thorac Cardiovasc Surg 2007; 134: 194 8.

25 Gandhi NR, Moll A, Sturm AW et al. Extensively drug-resistant

tuberculosis as a cause of death in patients co-infected with
tuberculosis and HIV in a rural area of South Africa. Lancet 2006; 368:
1575 80.

45 Wang HF, Lin HS, Jiang G. Pulmonary resection in the treatment of

multidrug-resistant tuberculosis: a retrospective study of 56 cases. Ann
Thorac Surg 2008; 86: 1640 5.

26 Pomerantz BJ, Cleveland JC Jr, Olson HK et al. Pulmonary resection for

multi-drug resistant tuberculosis. J Thorac Cardiovasc Surg 2001; 121:
44853.
27 Park SK, Lee CM, Heu JP et al. A retrospective study for the outcome of
pulmonary resection in 49 patients with multidrug-resistant tuberculosis.
Int J Tuberc Lung Dis 2002; 6: 143 9.
28 Chan ED, Laurel V, Strand MJ et al. Treatment and outcome analysis
of 205 patients with multidrug-resistant tuberculosis. Am J Respir Crit
Care Med 2004; 169: 1103 9.
29 Somocurcio JG, Sotomayor A, Shin S et al. Surgery for patients with
drug-resistant tuberculosis: report of 121 cases receiving communitybased treatment in Lima, Peru. Thorax 2007; 62: 41621.
30 World Health Organization. Guidelines for the Programmatic
Management of Drug-resistant Tuberculosis. Emergency Update 2008.
WHO/HTM/TB/2008.402. Geneva: World Health Organization, 2008.

46 Shiraishi YJ, Katsuragi N, Kita H et al. Aggressive surgical treatment of

multidrug-resistant tuberculosis. J Thorac Cardiovasc Surg 2009; 138:
1180 4.
47 Kang MW, Kim HK, Choi YS et al. Surgical treatment for
multidrug-resistant and extensive drug-resistant tuberculosis. Ann
Thorac Surg 2010; 89: 1597 602.
48 Sharma SK, Mohan A. Multidrug-resistant tuberculosis: a menace that
threatens to destabilize tuberculosis control. Chest 2006; 130: 162272.
49 Raviglione MC, Smith IM. XDR tuberculosisimplications for global
public health. New Engl J Med 2007; 356: 6569.
50 Mukherjee JS, Rich ML, Socci AR et al. Programmes and principles for
management of multidrug-resistant tuberculosis. Lancet 2004; 372:
47481.
51 Orenstein EW, Basu S, Sarita Shah N et al. Treatment outcomes
among patients with multidrug-resistant tuberculosis: systematic
review and meta-analysis. Lancet Infect Dis 2009; 9: 153 61.

31 Blumberg HM, Burman WJ, Chaisson RE et al. American Thoracic

Society/Centers for Disease Control and Prevention/Infectious Diseases
Society of America: treatment of tuberculosis. Am J Respir Crit Care Med
2003; 167: 60362.

52 Johnston JC, Shahidi NC, Sadatsafavi M et al. Treatment outcomes of

multidrug-resistant tuberculosis: a systematic review and meta-analysis.
PLoS ONE 4: e6914.

32 Moher D, Cook DJ, Eastwood S et al. Improving the quality of reports

of meta-analyses of randomized controlled trials: the QUOROM
statement. Lancet 1999; 354: 1896 900.

53 Torun T, Tahaoglu K, Ozmen I et al. The role of surgery and

fluoroquinolone in the treatment of multidrug resistant tuberculosis. Int
J Tuberc Lung Dis 2007; 11: 97985.

33 Smith TC, Spiegelhalter DJ, Thomas A. Bayesian approaches to

random-effects meta-analysis: a comparative study. Stat Med 1995;
14: 2685 99.

54 Lalloo UG, Naidoo R, Ambaram A. Recent advances in the medical

and surgical treatment of multi-drug resistant tuberculosis. Curr Opin
Pulm Med 2006; 12: 17985.

34 Gelman A, Carlin JB, Stern HS et al. Bayesian Data Analysis. Boca

Raton: CRC Press, 2003.

55 Wells CD, Cegielski JP, Nelson LJ et al. HIV infection and

multidrug-resistant tuberculosis: the perfect storm. J Infect Dis 2007;
196 (Suppl 1): 86 107.

35 Brooks S, Gelman A. General methods for monitoring convergence of

iterative simulations. J Comput Graph Stat 1998; 7: 43455.
36 Higgins JP, Thompson SG, Deeks JJ et al. Measuring inconsistency in
meta-analysis. BMJ 2003; 327: 55760.

56 Laserson KF, Thorpe LE, Leimane V et al. Speaking the same language:
treatment outcome definitions for multidrug-resistant tuberculosis. Int J
Tuberc Lung Dis 2005; 9: 6405.

37 Leuven MV, Groot MD, Shean KP et al. Pulmonary resection as an

adjunct in the treatment of multiple drug-resistant tuberculosis. Ann
Thorac Surg 1997; 63: 1368 73.

57 Tatsioni A, Zarin DA, Aronson N et al. Challenges in systematic

reviews of diagnostic technologies. Ann Intern Med 2005; 142:
1048 55.

38 Sung SW, Kang CH, Kim YT et al. Surgery increased the chance of cure
in multi-drug resistant pulmonary tuberculosis. Eur J Cardiothorac Surg
1999; 16: 18793.

58 WHO. The Stop TB Strategy: Building on and Enhancing DOTS to

Meet the TB-related Millennium Development Goals. Geneva: WHO,
2006.

1695

Downloaded from http://jac.oxfordjournals.org/ by guest on April 20, 2014

20 Leimane V, Riekstina V, Holtz TH et al. Clinical outcome of

individualised treatment of multidrug-resistant tuberculosis in Latvia: a
retrospective cohort study. Lancet 2005; 365: 31826.