| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Article |
From the Departments of Pathology and Thoracic
Surgery,*
National Defense Medical College,
Tokorozawa, Saitama, Japan; the Pathology
Section,
National Heart, Lung and Blood
Institute, National Institutes of Health, Bethesda, Maryland; and the
Department of Pulmonary and Mediastinal
Pathology,
Armed Forces Institute of
Pathology, Washington, District of Columbia
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Recent studies of the activity of MMPs and TIMPs in invasive neoplasms have indicated that these enzymes play important roles in the degradation of connective tissue that is associated with the development of metastases.4-8 In particular, MMP-2 and MMP-9 (gelatinase A and gelatinase B) are involved in the spread of carcinomas. These MMPs lyse partially degraded fibrillar collagens as well as elastin, but also have a high degree of affinity for type IV collagen, a major component of epithelial basement membranes, which are lysed during metastatic invasion. Galateau-Salle and colleagues9 conducted immunohistochemical studies of the reactivity of bronchial squamous preneoplastic and neoplastic lesions from cigarette smokers, including basal cell hyperplasia, squamous metaplasia, dysplasia, carcinoma in situ, and invasive squamous cell carcinoma, for MMPs, TIMPs, and type IV collagen. They concluded that bronchial squamous preneoplastic lesions show increased reactivity for MMPs, in association with destructive changes in the epithelial basement membranes, before the development of actual neoplasia or invasion of the subjacent connective tissue.
Most of the MMPs are produced in the form of biologically inactive proenzymes, which need to be activated to become biologically functional. Several mechanisms can result in the activation of MMP-2. The most important of these involves the action of a membrane-type 1 matrix metalloproteinase (MT-1-MMP), which cleaves a portion of the carboxy terminus of pro-MMP-2.10 The development of antibodies against MT-1-MMP has made it possible to evaluate the expression of both MMP-2 and its activating enzyme in the same tissue section, thus providing new information on the relationship between MMPs and the degradation of basement membranes.
Atypical adenomatous hyperplasia (AAH) was added to the group of preinvasive lesions in the 1999 WHO/IASLC classification, because it is thought to be a precursor of adenocarcinoma.1 It is a millimeter-sized nodular lesion, in which the alveoli and respiratory bronchioles are lined by slightly atypical pneumocytes.11,12 AAH is distinguished from nonmucinous BAC in that it has less cytological atypia, cellular crowding, and overlapping of nuclei, as well as less prominent nucleoli, and smaller cell size.13
The significance of central scarring in adenocarcinomas has been debated for several decades. Initially, it was thought that adenocarcinomas arose in association with pre-existing scars, and the concept of scar carcinoma was advanced.14,15 However, subsequent studies provided morphological,16-18 clinical, and collagen analyses that favored the concept that the scar in most lung carcinomas was caused by tumor invasion.19,20 The mechanisms proposed for this scarring include vascular17 or airway18 occlusion by tumor resulting in alveolar collapse and a dense fibrosing scar.
It has been reported that pulmonary adenocarcinomas with central scarring have a poorer prognosis than those without central scarring.3 However, only a few studies have compared the results of staining for type IV collagen and MMPs in AAH and in BACs versus the invasive components of pulmonary adenocarcinomas.7,8 In the present study, we have compared the preservation of the integrity of type IV collagen with the expression of MMPs and TIMPs, particularly of MMP-2 and MT-1-MMP, in both the lepidic and the invasive areas of pulmonary adenocarcinomas. We have also investigated the association between the expression of MMPs and the clinicopathological features and prognosis of these tumors.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The tissues studied consisted of 27 invasive adenocarcinomas with
lepidic areas and 5 AAH lesions, obtained by lobectomy or more
localized resection by thoracoscopy, without neoadjuvant chemotherapy
or radiotherapy. These procedures were performed at the National
Defense Medical College Hospital in Japan from 1989 to 1999. The study
was approved by the Committee on Human Research of the National Defense
Medical College. All lesions appeared to have originated from
peripheral areas of the lungs, and were fully resectable. The patients
were 10 men and 17 women and their mean age was 66 years (range, 44 to
81 years). The patients with AAH were three women and two men, with a
mean age of 65 years (range, 52 to 74 years). Clinical and pathological
data on these patients are summarized in Table 1
. The TNM classification was used for
clinical staging.21
The tumors were classified according
to the 1999 WHO classification.1
|
For histological study, tissues obtained from the largest cut surface of each tumor were fixed with buffered 10% formalin and embedded in paraffin. Sections (5 µm thick) were stained with hematoxylin and eosin (H&E). Each tumor was classified according to subtype (nonmucinous, mucinous, and mixed), growth pattern, and extent of lepidic growth. The extent of lepidic growth was estimated as a percentage of the overall tumor. The periodic acid-methenamine silver (PAM)-staining method was used to evaluate the basement membranes of the alveolar walls.22
Immunohistochemical Staining
Sections of paraffin-embedded tissues were used for the immunohistochemical staining procedures described below. The primary antibodies used in the study consisted of mouse monoclonal antibodies against the following components: MMP-1 (dilution, 1:500), MMP-3 (1:200), MMP-7 (1:100), MMP-9 (1:200), MT-1-MMP (1:100), MT-2-MMP (1:100), MT-3-MMP (1:100) (Chemicon International Inc., Temecula, CA), and type IV collagen (1:100) (DAKO, Carpinteria, CA) and rabbit polyclonal antibodies against MMP-2 (1:500), TIMP-1 (1:500) and TIMP-2 (1:100) prepared in the laboratory of Dr. W. Stetler-Stevenson (National Cancer Institute, National Institutes of Health, Bethesda, MD).
Sections were deparaffinized, rehydrated, treated with 0.4% pepsin (Sigma, St. Louis, MO) in 0.01 N HCl at 37°C (15 minutes for MMP-1, MMP-2, MMP-7, MT-1-MMP, and MT-2-MMP, and 30 minutes for MMP-9, TIMP-1, and TIMP-2). Pretreatment with 0.04% protease (P-5380, Sigma) was used for the demonstration of type IV collagen. No pretreatment was used to stain for MMP-3 and MT-3-MMP. The sections were treated with 0.3% hydrogen peroxide in methanol for 30 minutes at room temperature to block endogenous peroxidase activity. Then they were washed with phosphate-buffered saline (PBS) (0.01 mol/L, pH 7.2). The tissues to be reacted with monoclonal or polyclonal antibodies were incubated with 10% normal horse or goat serum, respectively, for 30 minutes to block nonspecific immunoglobulin binding. The sections were subsequently incubated for 2 hours at room temperature with the primary antibodies. After four washes with PBS (15 minutes each), the color was developed using the EnVision System (DAKO) and Vectastain kit (Vector Laboratories, Burlingame, CA) according to the manufacturers’ instructions. Then they were counterstained with hematoxylin and mounted. Negative immunohistochemical control procedures included: 1) omission of the primary antibody and 2) replacement of the primary antibody by normal mouse or rabbit IgG in appropriate concentrations. These control procedures gave negative results.
We observed both peripheral lepidic and central collapsed areas in the tumors examined. In both types of areas, the distribution of reactive tumor cells was graded as: 0, none; 1, 1 to 10%; 2, 10 to 50%; and 3, >50%. The intensity of the staining was graded as: 0, negative; 1, mild; 2, moderate; and 4, strong. Furthermore, we calculated the sum of these two parameters to evaluate the overall expression of MMPs and TIMPs, and created an immunohistochemical score based on this sum: 0, negative; 1 to 2, low; 3 to 4, moderate; and 5 to 6, high. Type IV collagen was rated as: + (normal) and - either partially (mainly in areas of lepidic growth) or completely (mainly in areas of central scars) destroyed.
Dual Labeling for Confocal Microscopy
For immunofluorescent staining, sections were deparaffinized, treated with pepsin or protease as described above, washed, and incubated with the mixture of 10% normal horse serum and 10% normal goat serum for 30 minutes at room temperature. They were then incubated overnight at 4°C with the mixture of a mouse monoclonal antibody (type IV collagen, MT-1-MMP, or TIMP-2) and a rabbit polyclonal antibody (MMP-2). In the mixture, the antibodies were diluted 1:20 for type IV collagen, 1:20 for MT-1-MMP, 1:20 for TIMP-2, and 1:100 for MMP-2. After washing with PBS, the sections were incubated with a mixture of the two secondary antibodies [fluorescein isothiocyanate-conjugated horse anti-mouse IgG, diluted 1:100 (FI-2000; Vector), and Texas red-conjugated goat anti-rabbit IgG, diluted 1:100 (TI-1000; Vector)] for 1 hour at room temperature. After washing, the nuclei were counterstained with 4'6-diamidino-2-phenylindole (DAPI-containing mounting medium H-1200; Vector), mounted, and examined with a confocal microscope (model TCS-4D/DMIRBE; Leica, Heidelberg, Germany) equipped with argon and argon-krypton lasers. In the preparations stained as described above, a green fluorescence indicated either type IV collagen, MT-1-MMP, or TIMP-2; a red fluorescence, MMP-2; a yellow fluorescence, co-localization of the red and green signals; and a blue fluorescence, nuclear DNA. A yellow autofluorescence, not indicative of co-localization, was observed in elastic fibers and in some macrophages and red blood cells. This autofluorescence was recognized by its presence in unstained sections and in immunohistochemical-negative control preparations.
Statistical Analysis
We compared the extent of lepidic growth and the preservation of type IV collagen with the expression of MMPs in the samples of adenocarcinoma with BAC component and AAH. The chi-square test was performed to determine the significance of the relationships between the expression of MMPs and the clinicopathological features, ie, tumor size, lymph node metastasis, lymphatic invasion, and vascular invasion. Survival was compared between patients whose tumors showed greater or equal to versus less than 50% lepidic areas. Survival rates were estimated by the method of Kaplan and Meier, and comparisons of these rates were made with the log-rank test. A P value of <0.05 was considered to be statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The TNM and staging information on the 27 patients in the study is
summarized in Table 1
. Thirteen tumors were stage 1a, 7 were stage 1b,
4 were stage 3a, and 1 each were stage 2a, 2b, and 3b. The total
follow-up time for all patients was 70.5 years with a mean for
individual patients of 2.3 years (range, 0.48 to 9.34 years). The
78-year-old female patient with a stage 1 tumor who died, developed
tumor recurrence in multiple bones and she died from tumor 12 months
after surgery.
Pathological Findings
The mean tumor size was 2.9 cm (range, 0.7 to 6.5 cm). Data on the
location of the tumors are summarized in Table 1
. The 27
adenocarcinomas were classified as the mixed subtype because they
showed mixtures of lepidic components and other subtypes of invasive
lesions. The lepidic component consisted of atypical pneumocytes that
proliferated along slightly thickened alveolar walls in a lepidic
manner. The mean extent of lepidic growth was 61% (range, 10 to 95%).
The invasive components consisted of varying combinations of the
acinar, papillary, and solid with mucin formation subtypes of
adenocarcinoma. In the invasive areas the epithelial proliferation was
associated with destruction of alveolar basement membranes and
proliferation of stromal cells (Figure 1, A and B)
. Inflammatory cells adjacent to the tumor cells and areas of
invasion of pulmonary lymphatics or blood vessels were recognized in 13
cases.
|
. Two distinct types of basement membranes (epithelial and
endothelial) were clearly recognized in lepidic areas of alveolar
walls in sections stained by the PAM method (Figure 3, A and B)
. In invasive areas, these
basement membranes were abnormally shaped or absent.
|
|
Type IV Collagen
In all tissue samples, the immunoreactivity for type IV collagen
was localized in the basement membranes of alveolar epithelial cells,
endothelial cells, and smooth muscle cells. There was a close
correspondence between the results obtained using the PAM stain and the
immunoperoxidase or immunofluorescence methods for type IV collagen.
Immunostaining for type IV collagen showed that the epithelial basement
membranes in areas of lepidic growth were not damaged in 16 cases
(Figure 3C)
but were partially destroyed in the other 11. There was a
negative correlation between the preservation of type IV collagen in
basement membranes and the occurrence of lymph node metastasis
(P = 0.03), as well as lymphatic
(P = 0.02) and vascular invasion
(P = 0.002). Furthermore, the 5-year survival
was lower in cases showing destruction of type IV collagen than in
those in which this component was well preserved in lepidic areas (20%
versus 100%, P = 0.004). The basement
membranes were either fragmented or completely destroyed in most
invasive areas (Figure 3D)
. Preservation of type IV collagen was
observed in all areas of AAH.
MMP-2
The reactivity for MMP-2 in the cytoplasm of the tumor cells
varied from negative to strong. This reactivity was mild to strong in
26 of the 27 cases of lepidic areas and mild to strong in all cases
having invasive areas (Figure 4, A and B)
. The staining for MMP-2 was significantly stronger in invasive areas
than in lepidic areas (P < 0.001). The MMP-2
score correlated with the occurrence of lymph node metastasis
(P = 0.01) and vascular invasion
(P = 0.003). In contrast, there was a negative
correlation between the expression of MMP-2 and the preservation of
type IV collagen (P = 0.02). Two of the five
cases of AAH showed high expression of MMP-2. The other three cases
were unreactive for this enzyme.
|
Reactivity for MT-1-MMP in areas of lepidic growth was observed in
23 of the 27 cases of adenocarcinoma (Figure 4C)
, but in 0 of the 5
cases of AAH. In invasive areas, this reactivity for MT-1-MMP was found
in 25 of the 27 patients (Figure 4D)
. The score of the reactivity for
MT-1-MMP was greater in invasive areas than in lepidic areas.
Other MMPs
The reactivity for other MMPs varied considerably in the cytoplasm of the tumor cells in both lepidic and invasive areas, ranging from negative to strong for MMP-3, MMP-9, MT-2-MMP, and MT-3-MMP, and from mild to strong for MMP-1 and MMP-7. The reaction for other MMPs in areas of AAH ranged from mild to moderate for MMP-3, MMP-7, MT-2-MMP, and MT-3-MMP, moderate to strong for MMP-1, and from negative to mild for MMP-9. No significant correlation was found between the expression of any of these MMPs and the clinicopathological features of the tumors.
TIMP-1 and TIMP-2
The staining for TIMP-1 (Figure 4, E and F)
and TIMP-2 (Figure 4, G and H)
ranged from negative to strong in lepidic and invasive areas.
The reaction of AAH cells was mild for TIMP-1 and ranged from negative
to mild for TIMP-2. The staining for TIMP-2 was significantly stronger
in invasive areas than in lepidic areas (P <
0.001). Furthermore, the expression of TIMP-2 correlated significantly
with that of MMP-2 in invasive areas (P =
0.046). No other correlations were found between these reactivities and
the clinicopathological features of the tumors. Data on the expression
of MMP-2, MT-1-MMP, TIMP-1, TIMP-2, and type IV collagen are summarized
in Table 2
.
|
The destruction of type IV collagen in lepidic and invasive areas
was readily recognized in sections subjected to dual staining for type
IV collagen and MMP-2 (Figure 5, A and B)
. In agreement with the results of the immunoperoxidase method, the
tumor cells in both lepidic and invasive areas were reactive for MMP-2.
Type IV collagen was preserved in lepidic areas of some cases in which
MMP-2 was strongly expressed. MMP-2 and MT-1-MMP were co-localized in
the cytoplasm of the tumor cells (Figure 5, C and D)
. Both MMP-2 and
TIMP-2 were strongly expressed in the cytoplasm of carcinoma cells in
invasive areas (Figure 5E)
. The expression of MMP-2, MT-1-MMP, and
TIMP-2 showed accentuation in the basal aspect of the carcinoma cells.
|
. No accentuation of the latter along
the basement membrane was observed. Correlation with Prognosis
The prognosis was significantly better in cases showing >50%
lepidic growth than in those showing <50% (P =
0.05) (Figure 6)
. The 5-year survival was
reduced in cases showing a moderate or high expression of MMP-2
compared with those showing negative or low expression of this enzyme
in lepidic areas (24% versus 100%, P =
0.008) (Figure 7)
.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Expression of MMP-2 has been observed in a variety of human tumors, and the importance of this enzyme in tumor cell invasion and metastasis has been widely recognized. Using a substrate capture enzyme-linked immunosorbent assay, Garbisa and colleagues6 found a good correlation of serum levels of MMP-2 with lung cancer metastasis and response to therapy. Kawano and colleagues7 reported that high levels of MMP-2 were expressed in most types of lung tumors. Kitamura and colleagues8 studied expression of MMP-2 in 48 peripherally located adenocarcinomas of the lung and 33 cases of AAH. They observed a positive reaction for MMP-2 in areas of lepidic growth in 17% of the cases and in invasive areas in 38%; the reaction was negative in AAH cases. In our study, some degree of reactivity for MMP-2 was observed in areas of lepidic growth in 26 cases (96.3%), in invasive areas of all 27 cases (100%) as well as in 2 of 5 cases of AAH. These results are comparable to those obtained by Galateau-Salle and colleagues,9 who found that MMP-2 is expressed in preneoplastic as well as in neoplastic bronchial squamous lesions.
Expression of MT-1-MMP, the cell-surface activator of pro-MMP-2, has been found in tumor cells in a variety of neoplasms, such as primary carcinomas of the stomach,23 pancreas,24 breast,25 thyroid,26 adrenal cortex,27 liver,28 and lung29 in association with activation of MMP-2. Tokuraku and colleagues30 reported that expression of MT-1-MMP correlated with activation of MMP-2 and with lymph node metastasis in carcinoma of the lung. However, we are not aware of studies of the expression of MT-1-MMP in other preneoplastic lesions.
We found no correlation between the expression of other MMPs (MMP-1, MMP-3, MMP-7, and MMP-9) and the clinicopathological features of the adenocarcinomas. Shima and colleagues31 reported that the tissue levels of MMP-3 correlated with vascular invasion and distant metastases in squamous cell carcinoma of the esophagus. Kossakowska and colleagues32 found that the levels of MMP-9 correlated with the histological grades of human malignant lymphomas. Using quantitative zymography, Davies and colleagues5 observed that levels of MMP-9 and activated MMP-2 correlated closely with tumor grade and invasiveness in transitional cell carcinoma of the bladder. These differences suggest that the expression of these markers may be influenced by the localization and histological type of tumor, the condition of adjacent normal tissues, and other factors.
TIMP-1 and TIMP-2 have been identified as protective factors for type IV collagen and fibrillar collagens. Kinoshita and colleagues33 reported that the addition of TIMP-2 inhibited the processing of pro-MMP-2 in a dose-dependent manner, and that TIMP-1 had only 10% or less of the inhibitory effect of TIMP-2. In our study, the expression of TIMP-1 was not found to correlate with the preservation of type IV collagen. The finding of increased TIMP2 expression in the invasive components of adenocarcinomas, compared to those in lepidic areas, provides evidence that the invading tumor cells promote the accumulation of extracellular matrix. This is consistent with the concept that the accumulation of extracellular matrix that results in the scar in pulmonary adenocarcinomas is promoted by TIMP-2 expression by the invasive tumor cells. It is possible that protective factors other than TIMP-2 also were associated with type IV collagen in such cases. Suzuki and colleagues34 and Yokose and colleagues35 recently highlighted the clinical importance of fibrous scars in pulmonary adenocarcinomas, in which they found a strong prognostic correlation with tumor scar size. They found a 100% 5-year survival in patients whose tumors had a scar size less than 5 mm whereas survival was significantly reduced in patients with larger scars. Thus, TIMP-2 expression by tumor cells and the development of pulmonary scars may relate to prognosis in pulmonary adenocarcinoma, even though in these sets of data we did not demonstrate a correlation between survival and the reactivity for TIMP-2.
Strong expression of MMP-2 was recognized in two of our five cases of AAH, but MT-1-MMP was not expressed in areas of AAH in any of these cases. The antibody that we used for the immunolocalization of MMP-2 does not distinguish between the proenzyme and the activated form of MMP-2. However, two observations suggest that the MMP-2 detected in our two cases of AAH was inactive: 1) the finding of an intact epithelial basement membrane in the areas of reactivity for MMP-2, and 2) the absence of MT-1-MMP from these areas. These findings are in contrast with those in invasive adenocarcinomas, in which we observed co-expression of MMP-2 and MT-1-MMP, frequently in association with damage to the type IV collagen in the basement membranes. Therefore, these observations suggest that the expression of MT-1-MMP is associated with progression of the preneoplastic lesion of AAH to adenocarcinoma. This progression would provide a potential mechanism for eventual invasion of the stroma by the neoplastic cells.
In conclusion, evaluation of the extent of the lepidic and invasive components of pulmonary adenocarcinoma, as demonstrated in histological sections by the PAM stain and in immunohistochemical preparations stained for type IV collagen and MMP-2, correlates with the prognosis in pulmonary adenocarcinomas. The development of scars in pulmonary adenocarcinoma is promoted by increased expression of TIMP-2 by the tumor cells as they grow invasively.
|
Footnotes |
|---|
Victor J. Ferrans is deceased.
Accepted for publication September 9, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  Y. Miyata, T. Iwata, K. Ohba, S. Kanda, M. Nishikido, and H. Kanetake Expression of Matrix Metalloproteinase-7 on Cancer Cells and Tissue Endothelial Cells in Renal Cell Carcinoma: Prognostic Implications and Clinical Significance for Invasion and Metastasis Clin. Cancer Res., December 1, 2006; 12(23): 6998 - 7003. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Dong, M. Sato, A. Sakurada, M. Sagawa, C. Endo, S. Wu, S. Yamanaka, A. Horii, and T. Kondo Computed tomographic images reflect the biologic behavior of small lung adenocarcinoma: They correlate with cell proliferation, microvascularization, cell adhesion, degradation of extracellular matrix, and K-ras mutation J. Thorac. Cardiovasc. Surg., September 1, 2005; 130(3): 733 - 739. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Ishikawa, K. Takenaka, K. Yanagihara, R. Miyahara, Y. Kawano, Y. Otake, S. Hasegawa, H. Wada, and F. Tanaka Matrix Metalloproteinase-2 Status in Stromal Fibroblasts, Not in Tumor Cells, Is a Significant Prognostic Factor in Non-Small-Cell Lung Cancer Clin. Cancer Res., October 1, 2004; 10(19): 6579 - 6585. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. S. Aljada, N. Ramnath, K. Donohue, S. Harvey, J. J. Brooks, S. M. Wiseman, T. Khoury, G. Loewen, H. K. Slocum, T. M. Anderson, et al. Upregulation of the Tissue Inhibitor of Metalloproteinase-1 Protein Is Associated With Progression of Human Non-Small-Cell Lung Cancer J. Clin. Oncol., August 15, 2004; 22(16): 3218 - 3229. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Pinto, P. E. d. O. Carvalho, L. Antonangelo, A. Garippo, A. G. P. da Silva, F. Soares, R. Younes, R. Beyruti, T. Takagaki, P. Saldiva, et al. Morphometric Evaluation of Tumor Matrix Metalloproteinase 9 Predicts Survival after Surgical Resection of Adenocarcinoma of the Lung Clin. Cancer Res., August 1, 2003; 9(8): 3098 - 3104. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Noda, S. Ishida, M. Inoue, K.-i. Obata, Y. Oguchi, Y. Okada, and E. Ikeda Production and Activation of Matrix Metalloproteinase-2 in Proliferative Diabetic Retinopathy Invest. Ophthalmol. Vis. Sci., May 1, 2003; 44(5): 2163 - 2170. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y Hakoda, Y Ito, A Nagate, K Minemura, K Utsumi, M Aoshima, and K Ohyashiki Increased collagenase activity in macrophages from bronchial lavage as a diagnostic marker of non-small cell lung cancer Thorax, February 1, 2003; 58(2): 122 - 126. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
