Annals of Oncology 17: 1673–1676, 2006
Published online 15 September 2006
Thyroid transcription factor 1—a new prognostic
factor in lung cancer: a meta-analysis
T. Berghmans1*, M. Paesmans2, C. Mascaux1, B. Martin1, A.-P. Meert1,
A. Haller3, J.-J. Lafitte4& J.-P. Sculier1
1Department of Intensive Care and Thoracic Oncology,2Data Centre,3Department of Pathology, Institut Jules Bordet, Centre des Tumeurs
de l’Universite ´ Libre de Bruxelles, Belgium;4Chest Department, Centre Hospitalier Universitaire Calmette, Lille, France
Received 3 April 2006; revised 30 June 2006; accepted 3 July 2006
Background: The aim of this study was to determine the prognostic role for survival of thyroid transcription factor 1
(TTF-1) in lung cancer.
Methods: Studies evaluating survival and TTF-1 in lung cancer patients, published until August 2005, were assessed
with a methodological scoring system. The required data for estimation of individual hazard ratios (HRs) for survival
were extracted from the publications and a combined HR was calculated.
Results: We identified 10 eligible papers, all dealing with non-small-cell lung cancer (NSCLC). Eight were meta-
analysed (evaluable studies). Seven studies included patients with local and/or locoregional diseases and three dealt
only with adenocarcinoma. Median methodological quality score was 65.9% (range = 31.8%–70.5%). TTF-1 positivity
was associated with statistically significant reduced or improved survival in one and four studies, respectively.
Combined HR for the eight evaluable studies was 0.64 [95% confidence interval (CI) = 0.41–1.00]. In the subgroup of
adenocarcinoma, the combined HR was 0.53 (95% CI = 0.29–0.95).
Conclusion: TTF-1 is a good prognostic factor for survival in NSCLC. Its effect appears also significant when the
analysis is restricted to patients with adenocarcinoma. This study supports the fact that TTF-1 could be included in
further prospective trials studying prognostic factors in NSCLC.
Key words: lung cancer, meta-analysis, prognostic factor, survival, systematic review, thyroid transcription factor 1
Non-small-cell lung cancer (NSCLC) is a major cause of death
in developed countries. Overall, survival is poor for the majority
of patients. Yet, in early stages, cure is obtained in <60% of the
cases . Advances have been obtained with neo-adjuvant
chemotherapy in surgical stages , with increased survival
associated with these treatments. Nevertheless, chemotherapy is
associated with significant morbidity and mortality needing for
a better definition of the patients who will most benefit from
such an approach. In this setting, prognostic factors could be of
potential value. Some clinical independent prognostic factors
for survival have already been identified: performance status
(PS), stage and, with lower impact, age, sex and weight loss [3,
4]. Some of these factors are useful when choosing the treatment
on an individual basis, principally disease stage, but for most of
the potential prognostic markers, their discriminant value is
insufficient to allow an individual decision.
In recent years, research has focused on the potential role of
new biological factors involved in the carcinogenic process, as
prognostic markers for survival in patients with lung cancer.
Results were frequently conflicting among the studies or
studies lacked statistical power. In order to answer this
question, the methodological group of the European Lung
Cancer Working Party (ELCWP) has carried out various
meta-analyses. We showed that vascular endothelial growth
factor , higher microvessel density , Epidermal
Growth Factor Receptor (EGFR) , HER-2/Neu , Ki-67 ,
K-Ras  and p53  are probable negative prognostic
factors, while Bcl-2 positivity  is associated with a better
The thyroid transcription factor 1 (TTF-1), also known as
Nkx2.1 or thyroid-specific enhancer-binding protein, is a 38-
kDa nuclear protein encoded by a gene located on chromosome
14q13. TTF-1 is expressed in the thyroid, the lung and the
diencephalum during embryogenesis. It plays a physiologic role
in the development and the morphogenesis of the thyroid and
the lung during embryogenesis. The role of TTF-1 in lung
carcinogenesis remains largely unexplained. It was suggested
that TTF-1 could promote cancerisation by regulating the
activity of proliferating cells and the formation of new vessels, at
least in adenocarcinoma , and by increasing the rate of cell
In lung cancer, TTF-1 is more frequently expressed in
adenocarcinoma and small-cell lung cancer than in squamous
*Correspondence to: Dr T. Berghmans, Institut Jules Bordet, rue He ´ger-Bordet,
1, B-1000 Brussels, Belgium. Tel: +32-2-541-31-91; Fax: +32-2-534-37-56;
ª 2006 European Society for Medical Oncology
by guest on June 13, 2013
cell carcinoma. The prognostic role of TTF-1 for survival in lung
cancer was assessed in few studies, mainly incorporating NSCLC
at local or locoregional stages. Their results are conflicting and
do not allow the drawing of definitive conclusions. The aim
of this study was to assess the possible role of TTF-1 as
prognostic factor for survival in lung cancer by carrying out a
systematic review of the literature followed by a meta-analysis.
material and methods
To be eligible for the systematic review, a study had to fulfil the following
criteria: to deal with lung cancer only (any stage or histology), to assess
the relationship between TTF-1 and survival and to have been published as
a full paper in the English, Dutch or French language. Abstracts were
excluded due to insufficient data to evaluate the methodological quality of
the trial and/or to carry out meta-analysis. Studies were identified by an
electronic search on Medline data bank using the following key words:
‘lung neoplasms’, ‘TTF-1’ and ‘thyroid transcription factor 1’. The
references reported in all the identified studies were used to complete this
search which ended in August 2005.
To assess the methodological quality of the publications, each study was
read independently by seven investigators according to the ELCWP scale
previously published  and applied in other meta-analyses [5–10, 12].
Each item was assessed using an ordinal scale (possible values = 2, 1, 0). As
the assessed items were objective ones, a consensus was always obtained on
each value. All items were separated into four categories, each scored on
a maximum of 10 points: the scientific design, the description of the
laboratory methods, the generalisability of the results and the analysis of the
study data. The final scores were expressed as percentages, ranging from 0%
to 100%, higher values reflecting better methodological quality. Studies
included in the systematic review were called ‘eligible’ and those providing
sufficient data for the meta-analysis ‘evaluable’.
A study was considered as significant if the P value for the statistical test
comparing survival distributions between the groups with and without
TTF-1 expression was <0.05. A study was called ‘positive’ or ‘negative’ when
TTF-1 positivity was identified as a significant favourable or unfavourable
prognostic factor for survival, respectively. These studies were further called
‘significant’ones. Finally, a study was called ‘not significant’ if no statistically
significant difference between the two groups was detected.
The association between two continuous variables was measured by the
Spearman’s rank correlation coefficient. Non-parametric tests were used to
compare the distribution of the quality scores according to the value of
a discrete variable (Mann–Whitney tests for dichotomic variables or
Kruskal–Wallis tests for multiple class variables).
For the quantitative aggregation of the survival results, we measured the
impact of TTF-1 positivity on survival by hazard ratio (HR) between the two
survival distributions. For each trial, this HR was estimated by a method
depending on the data provided in the publication. The most accurate
method consisted of calculating the estimated HR and its standard error
from the reported results or to calculate them directly using two of the
following parameters: the O ? E statistic (difference between numbers of
observed and expected events), the confidence interval (CI) for the HR and
the log-rank statistic or its P value. If these were not available, the total
numbers of events, the number of patients at risk in each group and the log-
estimate. Finally, if the only exploitable data were in the form of graphical
representations of the survival distributions, survival rates at some specified
with the assumption that the rate of patients censored was constant during
the study follow-up . If this last method was used, three independent
persons read the curves to reduce the inaccuracy in the extracted survival
rates. The individual HR estimates were combined into an overall HR using
Peto’s method . If homogeneity among the studies was confirmed by
carrying out a chi-square testfor heterogeneity, we useda fixed-effect model;
otherwise, we used a random-effect model. By convention, an observed HR
<1 implied a better survival for the group with TTF-1 expression and was
considered statistically significant if the 95% CI did not overlap 1.
Ten publications, published between 1999 and 2005, were
eligible for the systematic review [13, 14, 17–24]. All these
publications concerned different cohorts of patients. The total
number of included patients was 1101, ranging from 50 to 284
patients per study (median = 93). The main characteristics of
the 10 eligible publications are reported in Table 1. TTF-1
expression was always detected by immunohistochemistry. The
same mAb was used except by Puglisi et al. . There was
a doubt for Shah et al.  who used an antibody furnished by
Dako (Carpinteria, CA) which was the producer of the majority
of mouse mAb 8G7G3/1 employed by the other authors. Two
studies were not included in the meta-analysis due to a lack of
data in the publications, not allowing the HR calculation.
Half of the studies were significant (all evaluable for meta-
analysis) of which only one identified TTF-1 as a poor
prognostic factor for survival; four studies demonstrated that
TTF-1 positivity is associated with better survival. The five
others showed no statistically significant impact of TTF-1 on
survival (three evaluable for meta-analysis).
The overall quality score for the 10 eligible studies ranged
from 31.8% to 70.5% with a median of 65.9%. The individual
results of the quality scores are reported in Table 1.
The individual HR of the eight evaluable studies (652 patients)
[13, 14, 18–22, 24] was calculated as follows: three studies
statistic; forthe threeremaining ones,HRhad tobeextrapolated
from the graphical representation of the survival distributions.
The test for heterogeneity including the eight evaluable studies
was highly significant (P <0.001). We used a random-effect
model in calculating the overall HR which was 0.64 (95% CI =
0.41–1.00) (Figure 1). The heterogeneity was mainly attributed
to the study of Puglisi et al. . We secondly carried out the
same meta-analysis without this publication (564 patients). The
HR (random model) was then 0.56 (95% CI = 0.34–0.90).
According to the literature, TTF-1 is mainly expressed in
adenocarcinoma. We looked at the prognostic role of TTF-1 in
this subgroup of patients. Four studies (281 patients) included
adenocarcinomas only [13, 18, 21] or reported separate data for
those patients . The combined HR for the four evaluable
studies (random-effect model) was 0.53 (95% CI = 0.29–0.95).
We carried out a systematic review of the French and English
language literature todetermine the impact of TTF-1 on survival
Annals of Oncology
1674 | Berghmans et al.Volume 17|No. 11| November 2006
by guest on June 13, 2013
in patients with lung cancer. We observe a significant role of
TTF-1 in NSCLC and in the subgroup of adenocarcinomas.
TTF-1 should be considered as a good prognostic factor for
survival in NSCLC.
Prognostic factors have been extensively studied in lung
cancer . Some clinical and biological parameters have been
identified, mainly stage and PS [3, 4]. Some of them have
a direct impact on the therapeutic strategy, essentially disease
stage. However, with similar disease extent, prognosis could be
dissimilar. Stages are not homogeneous and it is difficult, with
the currently available prognostic factors, to determine which
patients will survive and who will die despite identical
treatments. There is a need for new prognostic factors. In this
way, recent knowledge in the lung cancer biology prompted
lung cancer specialists to assess the prognostic role of new
biological factors in this population. Unfortunately, these are
frequently small size studies with limited statistical power.
Therefore, meta-analyses are needed to answer the question, as
our group published these last years [5–12].
TTF-1 is an interesting marker, currently used in routine
clinical practice to distinguish lung adenocarcinoma from
adenocarcinoma metastatic to the lung. Moreover, TTF-1 is also
expressed in lung squamous carcinoma. The role of TTF-1 in
lung carcinogenesis remains unclear, although different
explanations are proposed [13, 14]. Ten studies assessing the
relationship between TTF-1 and survival (Table 1) have been so
far published [13, 14, 17–24]. Four demonstrated a positive
impact of TTF-1 on survival, one observed a detrimental effect
of TTF-1 and in the last five studies, no survival difference was
found between TTF-1-positive and -negative tumours. Once
more, the conflicting results of the literature with underpowered
studies require carrying out a quantitative aggregation of the
individual results (meta-analysis).
We found that TTF-1 positivity was associated with better
survival in NSCLC. This effect is essentially demonstrated in
early and locally advanced diseases (stages I–III) as there is only
one study incorporating stage IV NSCLC without separate
results in this group of patients. Also, the positive impact of
TTF-1 is demonstrated in adenocarcinoma but the lack of data
in other NSCLC histologies does not allow the drawing of
definite conclusions. Randomised trials of neo-adjuvant
chemotherapy demonstrated that an approach combining
surgery and medical treatment increased absolute cure rates by
5%–10%. However, all the patients do not benefit from these
treatments, needing to better delineate those who need to be
treated. In this setting, it could be hypothesised that new
biological factors are of potential usefulness. On the basis of our
results, TTF-1 could be selected for a prospective study with
multivariate analysis including the most known prognostic
factors in NSCLC such as stage and PS.
Our analysis had to deal with problems of heterogeneity.
There was a significant heterogeneity among the eight evaluable
Table 1. Main characteristics and results of 10 eligible studies assessing the prognostic role of TTF-1 on survival in patients with lung cancer
First author Year HistologyStageN Percentage of
Puglisi 1999 NSCLCI–III8838 Antiserum raised
HR and 95% CI
Log-rank + n events
HR and 95% CI
Log-rank + n events
N, number of patients; TTF-1, thyroid transcription factor 1; mAb, mAb used for TTF-1 detection by immunohistochemistry; HR, hazard ratio; HR
estimation, description of the methods used to estimate the individual HR according to the three different methods described in the statistical section; QS,
quality score; NSCLC, non-small-cell lung cancer; ADC, adenocarcinoma; NS, not significant; BAC, bronchioalveolar carcinoma; CI, confidence interval.
Figure 1. Meta-analysis of the eight evaluable studies assessing thyroid
transcription factor 1 (TTF-1) in non-small-cell lung cancer (NSCLC).
Hazard ratio (HR) and 95% confidence limits (CL) for survival
comparison in studies evaluating TTF-1 status in NSCLC. HR <1 implies
a survival advantage for the group with TTF-1 expression. The square size
is proportional to the number of patients included in the study. The centre
of the diamond-shaped lozenge at the bottom of the figure gives the
combined HR of the meta-analysis and its extremities, the 95% CI.
HR = 0.64; 95% CI = 0.41–1.00. Total number of patients = 652.
Annals of Oncology
Volume 17 | No. 11 | November 2006 doi:10.1093/annonc/mdl287 | 1675
by guest on June 13, 2013
studies included in the meta-analysis. Heterogeneity is
principally related to the study of Puglisi et al.  in which the
results were opposite to those found in the other publications.
One explanation could be related to the immunohistochemistry
techniques. Most importantly, all the authors used the same
mouse mAb (8G7G3/1) except Puglisi et al. , who employed
an antiserum raised in rabbits. The exclusion of this study from
the analysis decreased only partially the heterogeneity for which
there is different explanations either related to the
immunohistochemistry techniques (various definitions of
threshold positivity, use of the mAb at different concentrations
and dissimilar staining protocols) or related to the patients’
characteristics (type of patients, disease characteristics). More
precisely, the cut-off definition of TTF-1 positivity was at least
one positive cells, >5% or >10% in two, three and one studies,
respectively. A complex score was used in two other studies and
no clear definition was used in the last two studies.
Some other biases could have occurred. We carried out
a methodological assessment of the studies to avoid some
selection biases, as we carried out in prior studies about
significant and non-significant and between evaluable and non-
evaluable studies. Nevertheless, the limited number of studies in
each group did not allow us to perform meaningful statistical
comparisons. However, this approach does not prevent all
fully published studies only, method of extrapolation of HR,
validity of a meta-analysis based on systematic review of the
literature as compared with those based on individual data
were already discussed in our previous papers [5–12].
In conclusion, we observe that TTF-1 positivity is associated
with better survival in NSCLC, mainly in early and locally
advanced stages and in adenocarcinomas. Our present results
and those from previous meta-analyses carried out by our group
could serve to set up a prospective trial aiming at better
delineating the patients who will most benefit from neo-
adjuvant chemotherapy in surgical stage lung cancers. This trial
must include, in a multivariate analysis, the classical well-
defined prognostic factors in lung cancer and the new biological
prognostic markers that have been identified.
CM is supported by a fellowship from the National Fund for
1. Mountain CF. Revisions in the International System for Staging Lung Cancer.
Chest 1997; 111: 1710–1717.
2. Berghmans T, Paesmans M, Meert AP et al. Survival improvement in resectable
non-small cell lung cancer with (neo)adjuvant chemotherapy: results of a meta-
analysis of the literature. Lung Cancer 2005; 49: 13–23.
3. Paesmans M, Sculier JP, Libert P et al. Prognostic factors for survival in
advanced non-small-cell lung cancer: univariate and multivariate analyses
including recursive partitioning and amalgamation algorithms in 1,052 patients.
The European Lung Cancer Working Party. J Clin Oncol 1995; 13: 1221–1230.
4. Strauss GM. Prognostic markers in resectable non-small cell lung cancer.
Hematol Oncol Clin North Am 1997; 11: 409–434.
5. Delmotte P, Martin B, Paesmans M et al. VEGF et survie des patients atteints
de cancer pulmonaire: revue syste ´matique de la litte ´rature et me ´ta-analyse.
Rev Mal Respir 2002; 19: 577–584.
6. Meert AP, Paesmans M, Martin B et al. The role of microvessel density on
the survival of patients with lung cancer: a systematic review of the literature
with meta-analysis. Br J Cancer 2002; 87: 694–701.
7. Meert AP, Martin B, Delmotte P et al. The role of EGF-R expression on patient
survival in lung cancer: a systematic review with meta-analysis. Eur Respir J
2002; 20: 975–981.
8. Meert AP, Martin B, Paesmans M et al. The role of HER-2/neu expression on
the survival of patients with lung cancer: a systematic review of the literature.
Br J Cancer 2003; 89: 959–965.
9. Martin B, Paesmans M, Mascaux C et al. Ki-67 expression and patients survival
in lung cancer: systematic review of the literature with meta-analysis. Br J
Cancer 2004; 91: 2018–2025.
10. Mascaux C, Iannino N, Martin B et al. The role of RAS oncogene in survival
of patients with lung cancer: a systematic review of the literature with
meta-analysis. Br J Cancer 2005; 92: 131–139.
11. Steels E, Paesmans M, Berghmans T et al. Role of p53 as a prognostic factor for
survival in lung cancer: a systematic review of the literature with a meta-analysis.
Eur Respir J 2001; 18: 705–719.
12. Martin B, Paesmans M, Berghmans T et al. Role of Bcl-2 as a prognostic factor
for survival in lung cancer: a systematic review of the literature with meta-
analysis. Br J Cancer 2003; 89: 55–64.
13. Pelosi G, Fraggetta F, Pasini F et al. Immunoreactivity for thyroid transcription
factor-1 in stage I non-small cell carcinomas of the lung. Am J Surg Pathol 2001;
14. Puglisi F, Barbone F, Damante G et al. Prognostic value of thyroid transcription
factor-1 in primary, resected, non-small cell lung carcinoma. Mod Pathol 1999;
15. Parmar MK, Torri V, Stewart L. Extracting summary statistics to perform meta-
analyses of the published literature for survival endpoints. Stat Med 1998; 17:
16. Yusuf S, Peto R, Lewis J et al. Beta blockade during and after myocardial
infarction: an overview of the randomized trials. Prog Cardiovasc Dis 1985;
17. Au NH, Cheang M, Huntsman DG et al. Evaluation of immunohistochemical
markers in non-small cell lung cancer by unsupervised hierarchical clustering
analysis: a tissue microarray study of 284 cases and 18 markers. J Pathol 2004;
18. Barlesi F, Pinot D, Legoffic A et al. Positive thyroid transcription factor 1 staining
strongly correlates with survival of patients with adenocarcinoma of the lung.
Br J Cancer 2005; 93: 450–452.
19. Haque AK, Syed S, Lele SM et al. Immunohistochemical study of thyroid
transcription factor-1 and HER2/neu in non-small cell lung cancer: strong thyroid
transcription factor-1 expression predicts better survival. Appl Immunohistochem
Mol Morphol 2002; 10: 103–109.
20. Myong NH. Thyroid transcription factor-1 (TTF-1) expression in human
lung carcinomas: its prognostic implication and relationship with
expressions of p53 and Ki-67 proteins. J Korean Med Sci 2003;18: 494–500.
21. Saad RS, Liu YL, Han H et al. Prognostic significance of thyroid transcription
factor-1 expression in both early-stage conventional adenocarcinoma and
bronchioloalveolar carcinoma of the lung. Hum Pathol 2004; 35: 3–7.
22. Shah L, Walter KL, Borczuk AC et al. Expression of syndecan-1 and expression
of epidermal growth factor receptor are associated with survival in patients
with nonsmall cell lung carcinoma. Cancer 2004; 101: 1632–1638.
23. Stenhouse G, Fyfe N, King G et al. Thyroid transcription factor 1 in pulmonary
adenocarcinoma. J Clin Pathol 2004; 57: 383–387.
24. Tan D, Li Q, Deeb G et al. Thyroid transcription factor-1 expression prevalence
and its clinical implications in non-small cell lung cancer: a high-throughput
tissue microarray and immunohistochemistry study. Hum Pathol 2003; 34:
25. Buccheri G, Ferrigno D. Prognostic factors in lung cancer: tables and comments.
Eur Respir J 1994; 7: 1350–1364.
Annals of Oncology
1676 | Berghmans et al.Volume 17|No. 11| November 2006
by guest on June 13, 2013