Eur J Cardiothorac Surg 2004;26:1211-1215
© 2004 Elsevier Science NL
Lung adenocarcinoma can be subtyped according to tumor dimension by computed tomography mediastinal-window setting. Additional size criteria for clinical T1 adenocarcinoma
Yukinori Sakaoa,*,
Takahiko Nakazonob,
Shinji Tomimitsua,
Yuji Takedaa,
Toru Sakuragia,
Masafumi Natsuakia,
Tsuyoshi Itoha
a Department of Thoracic and Cardiovascular Surgery, Saga Medical School, Nabesima 5-1-1, Saga 849-8501, Japan
b Department of Radiology, Saga Medical School, Nabesima 5-1-1, Saga 849-8501, Japan
Received 31 May 2004;
received in revised form 19 August 2004;
accepted 23 August 2004.
* Corresponding author. Tel.: +81 952 34 2345; fax: +81 952 34 2061. (E-mail: sakao{at}post.saga-med.ac.jp).
 |
Abstract
|
|---|
Objective: We sought to verify that the size of the solid component, which can be evaluated using the computed tomography mediastinal-window setting, provides new criteria for CT classification of lung adenocarcinoma. Methods: Between 1994 and September 2002, we examined 60 patients who were clinically classified with stage T1 adenocarcinoma of the lung and normal serum CEA, who underwent standard surgical procedures. Tumor maximum dimension was evaluated using two different CT-imaging settings: the lung window (lDmax), and the mediastinal window (mDmax). We analyzed the relationships between prognosis or lymph node involvement and tumor dimensions. Results: The mDmax was a significant (OR 1.11, P=0.02) predictive factor for lymph node metastasis. However, lDmax was not significant (P=0.83). Age, gender, lDmax, mDmax, and lymph node involvement were analyzed as predictive factors for prognosis. In univariate analysis, mDmax and lymph node involvement were significant predictive factors for prognosis (OR 1.07, P=0.01; OR 2.56, P=0.04; respectively). In multivariate analysis, mDmax was a significant predictive factor for prognosis (OR 1.06, P=0.04). We then classified the C-T1 adenocarcinoma patients into three groups according to mDmax: T1a (
10mm), T1b (from 11 to 20mm), and T1c (from 21 to 30mm). There was a significant difference between the three groups: the disease-free 5-year survivals were 93.3, 58.1, and 32.7%, respectively (P=0.01). Conclusions: The mDmax can give additional, useful prognostic data. This finding may provide new criteria for CT classification of lung adenocarcinoma.
|
1. Introduction
|
|---|
Recently, improvements in diagnostic radiology have introduced the concept that the degree of dominance of the ground-glass attenuation (GGA) component to the non-GGA component (‘solid component’) is associated with prognosis in small adenocarcinoma. That is, the GGA-dominant type adenocarcinoma has shown better prognosis than has the non-GGA-dominant type [1–8]. Most of the GGA component on CT, which disappears in the mediastinal window, is related to a bronchioloalveolar lesion that is thought to be a non-invasive cancer; conversely, the solid component on CT, which does not disappear in the mediastinal window, is related to non-replacing tumor growth or to a large component of fibrous scar, which has been described as indicative of the invasiveness of the tumor [1–4,9–12].
In consideration of previous studies [1–11], we hypothesized that the size of the solid component of the tumor is a more important predictive factor for prognosis than the total tumor size, which includes the GGA lesion.
In this retrospective study, we have tried to demonstrate that the tumor dimension as evaluated using the mediastinal-window setting, which is able to indicate the size of the solid lesion, provides an important, additional predictive factor. This may provide us with new criteria for CT classification of lung adenocarcinoma, and aid in determining the most appropriate therapeutic strategy.
|
2. Patients and methods
|
|---|
Between 1994 and September 2002, 451 patients underwent surgical resection of primary lung cancer, within our department. Of those, 60 cases that had been classified clinically as having T1 adenocarcinoma of the lung, with normal serum CEA, and treated with standard surgical procedures (lobectomy+hilar and mediastinal node dissection without neoadjuvant therapy), were studied. The thin-section chest CT images (CT with 2.0 or 3.0mm section thickness, for imaging the tumor) and the clinicopathological records of the patients were examined. Tumor dimension was evaluated using a pair of different CT-imaging conditions: the lung window (level –500HU, width 1500HU) and the mediastinal window (level 60, width 350HU) settings. The CT scan films were independently examined by two authors (Y.S. and S.T.) to evaluate the maximum dimension of the tumor on both the lung (lDmax) and mediastinal (mDmax) windows (Fig. 1) [1]. The relationships between prognosis or lymph node metastasis and the tumor dimension evaluated under the two conditions were analyzed.
View larger version (107K):
[in this window]
[in a new window]
|
Fig. 1. Tumor dimension evaluated using a pair of different CT-imaging conditions: the lung window (mDmax) setting, with level –500HU and width 1500HU, and the mediastinal window (mDmax) setting, with level 60 and width 350HU.
|
|
The patient characteristics are summarized in Table 1. In all patients, pre-operative staging was assessed according to the TNM classification of the International Union Against Cancer [13] using chest CT, abdominal CT or ultrasonography, brain CT or magnetic resonance imaging, and bone scanning. Clinical mediastinal and hilar lymph node status was assessed as positive if the results of the chest CT showed that the shorter axis was larger than 1.0cm. The follow-up period ranged from 12 to 106 months (median 49 months). We excluded tumors of type 100% GGA from this study because they were believed, on microscopic examination, to be either non-invasive or pre-cancerous lesions [5–12]. We excluded mucinous BAC for the same reason. Furthermore, we excluded patients with a serum CEA level higher than 5ng/ml (based on the 95% specificity level for benign lung disease in our institute); this was because increased serum CEA has been found to be a predictor for poor prognosis after surgery, independent of tumor size or lymph node metastasis—especially in early stage lung cancer [1–4].
|
3. Statistical analysis
|
|---|
Disease-free survival rate was used to quantify prognosis in this study. The disease-free survival rates were calculated using the Kaplan–Meier method. Univariate analyses were performed using the logistic regression procedure or Cox proportional hazard model. Multivariate analyses were performed by means of the Cox proportional hazard model when the P-value was less than 0.2 in univariate analysis. Values of the variables were expressed as mean±SD. All of the statistical analyses were performed using StatView J 5.0 (SAS Institute Inc, Cary, NC). A P-value of less than 0.05 was treated as significant.
|
4. Results
|
|---|
4.1. Lymph node metastasis and tumor size
The mDmax was a significant (OR 1.11, P=0.02) predictive factor for lymph node metastasis in C-T1 adenocarcinoma. However, lDmax was not a significant (OR 1.01, P=0.83) predictive factor for lymph node metastasis (Table 2).
4.2. Predictive factors for disease free survival
Age, gender, lDmax, mDmax, and lymph node involvement (positive vs. negative) were analyzed as predictive factors for prognosis.
In univariate analysis, mDmax and lymph node involvement were significant predictive factors for prognosis (mDmax: OR 1.07, P=0.01; lymph node involvement: OR 2.56, P=0.04) (Table 3). In multivariate analysis (of factors with a P-value of less than 0.2 in univariate analysis), mDmax was a significant predictive factor for prognosis (OR 1.06, P=0.04); however, lymph node involvement was not significant (OR 1.57, P=0.36) (Table 4).
4.3. Subgroup according to mDmax
We tried classifying the C-T1 adenocarcinoma patients (with normal serum CEA) into three groups according to mDmax: T1a (N=15, mDmax less than or equal to 10mm; T1b (N=24, mDmax from 11 to 20mm); and T1c (N=21, mDmax from 21 to 30mm) (Fig. 2). There was a significant difference between the three groups; the disease-free 5-year survivals were 93.3, 58.1, and 32.7%, respectively (P=0.01). In addition, we examined the association between the subgroups and the incidence of lymph node metastasis. The incidence of lymph node involvement was higher in the larger-size subgroups (T1a, N2:1; T1b, N1:1 and N2:3; T1c, N1:2 and N2:4), but was not significant (P=0.22) (Fig. 3). Even if we excluded the node positive patients in the analysis of survival, the result was almost same (100% in T1a, 64.4% in T1b, and 34.2% in T1c) (Fig. 4).
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 2. The 5-year disease-free survival rate in patients diagnosed with clinical T1 adenocarcinoma of the lung depended on tumor maximum diameter as evaluated using the mediastinal window (mDmax) on CT. T1a: mDmax less than or equal to 10mm; T1b: mDmax from 11 to 20mm; T1c: mDmax from 21 to 30mm.
|
|
View larger version (62K):
[in this window]
[in a new window]
|
Fig. 3. Incidence of lymph node involvement depended on tumor maximum diameter as evaluated using the mediastinal window (mDmax) on CT. T1a: mDmax less than or equal to 10mm; T1b: mDmax from 11 to 20mm; T1c: mDmax from 21 to 30mm.
|
|
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 4. The 5-year disease-free survival rate in patients diagnosed with clinical T1 and pN0 adenocarcinoma of the lung depended on tumor maximum diameter as evaluated using the mediastinal window (mDmax) on CT. T1a: mDmax less than or equal to 10mm; T1b: mDmax from 11 to 20mm; T1c: mDmax from 21 to 30mm.
|
|
|
5. Comment
|
|---|
In this study, we discovered that tumor dimension as determined from the mediastinal window (mDmax) gives additional, useful prognostic data that cannot be measured by conventional staging modalities. Specifically, mDmax was an important predictive factor not only for lymph node involvement but also for prognosis—independent of lymph node metastasis—in T1 lung adenocarcinoma.
Furthermore, we showed that C-T1 adenocarcinoma, when classified into three groups according to mDmax, is highly correlated with prognosis as long as serum CEA is within the normal range. When mDmax was larger than 20mm, the 5-year disease-free survival was 32.7% overall, for C-T1, and 34.2% even for C-T1-pN0 adenocarcinoma. However, when mDmax was smaller than 10mm, the 5-year disease-free survival was better than 90%. These findings should provide us with new criteria for CT classification of lung adenocarcinoma.
Recently, there have been many reports on the correlation between CT findings and pathological features in lung adenocarcinoma [2,5–8]. The GGA component is most often recognized as a bronchioloalveolar carcinoma (BAC) component that is defined as a non-invasive lesion [5–8,12]. However, the solid component is recognized as either a fibrous scar or a component of non-BAC adenocarcinoma (non-replacing type adenocarcinoma) with fibrous scar [2]. It has been reported that the larger or more desmoplastic fibrous scar, which is described as indicative of the invasiveness of the tumor, is associated with poorer prognosis in clinicopathological analysis [9–11].
It is suggested that mDmax indicates the degree of invasiveness of the tumor and that it can predict the prognosis. In support of this inference, Okada et al. reported a correlation between dominance of the solid component and high incidence of vascular or lymphatic invasions [3].
The role of lDmax as a predictor for prognosis has been unclear. It has been reported that patients with adenocarcinoma tumor size larger than 20mm show poorer outcome than patients with C-T1 or C-IA [1–4]. However, once lDmax is analyzed with respect to the parameter that is related to mDmax (i.e. the tumor shadow disappearance rate: TDR) in multivariate analysis, it is found to be a non-related factor for prognosis [1,3]. It is suggested that, although lDmax may indeed be a predictive prognostic factor, mDmax is a more important one.
We excluded patients with a serum CEA level higher than 5ng/ml because serum CEA concentration has been revealed to be an important predictive factor for prognosis after surgery, independent of tumor size or lymph node metastasis, especially in clinical early-stage lung cancer [2–4]. When such patients were included in the study, the difference in prognosis between T1b and T1c became smaller (data not shown). Thus, in C-T1 adenocarcinoma, mDmax is highly correlated with prognosis as long as serum CEA is within the normal range.
There were a few instances of tumor diameter (lDmax) less than 10mm, in the present study, because most of such cases underwent limited surgery (with limited or no lymph node dissection). Therefore, very small lesions (lDmax less than 10mm) should be analyzed in a separate series.
In summary, mDmax is presented as one of the simplest and most reliable variables for predicting prognosis in C-T1 lung adenocarcinoma. This finding should provide new criteria for CT classification of lung adenocarcinoma and for selection of the most efficacious therapeutic strategy.
|
Acknowledgments
|
|---|
We thank Dr Edmund J. Miller, Chief, Surgical Immunology, North Shore University Hospital, for critical reviews.
|
References
|
|---|
- Takamochi K, Nagai K, Yoshida J, Suzuki K, Ohde Y, Mishimura M, Sasaki S, Nishiwaki Y. Pathologic N0 status in pulmonary adenocarcinoma is predictable by combining serum carcinoembryonic antigen level and computed tomographic findings. J Thorac Cardiovasc Surg 2001;122:325-330.[Abstract/Free Full Text]
- Sakao Y, Nakazono T, Sakuragi T, Natsuaki M, Itoh T. Predictive factors for survival in surgically resected clinical IA peripheral adenocarcinoma of the lung. Ann Thorac Surg 2004;77:1157-1162.[Abstract/Free Full Text]
- Okada M, Nishi W, Sakamoto T, Uchino K, Tsubota N. Discrepancy of computed tomographic image between lung and mediastinal windows as a prognostic implication in small lung adenocarcinoma. Ann Thorac Surg 2003;76:1828-1832.[Abstract/Free Full Text]
- Sakao Y, Sakuragi T, Natsuaki M, Itoh T. Clinicopathological analysis of prognostic factors in clinical IA peripheral adenocarcinoma of the lung. Ann Thorac Surg 2003;75:1113-1117.[Abstract/Free Full Text]
- Aoki T, Nakata H, Watanabe H, Nakamura K, Kasai T, Hashimoto H, Yasumoto T, Kido M. Evolution of peripheral lung adenocarcinomas: CT findings correlated with histology and tumor doubling time. Am J Roentgenol 2000;174:763-768.[Abstract/Free Full Text]
- Aoki T, Tomoda Y, Watanabe H, Nakata H, Kasai T, Hashimoto H, Kodate M, Osaki T, Yasumoto K. Peripheral lung adenocarcinoma: correlation of thin-section CT findings with histologic prognostic factors and survival. Radiology 2001;220:803-809.[Abstract/Free Full Text]
- Kondo T, Yamada K, Noda K, Nakayama H, Kameda K. Radiologic–prognostic correlation in patients with small pulmonary adenocarcinomas. Lung Cancer 2002;36:49-57.[CrossRef][Medline]
- Takashima S, Maruyama Y, Hasegawa M, Yamada T, Honda T, Kadoya M, Sone S. Prognostic significance of high-resolution CT findings in small peripheral adenocarcinoma of the lung: a retrospective study on 64 patients. Lung Cancer 2002;36:289-295.[CrossRef][Medline]
- Noguchi M, Morikawa A, Kawasaki M, Matsuno Y, Yamada T, Hirohashi E. Small adenocarcinoma of the lung: histologic characteristics and prognosis. Cancer 1995;75:2844-2852.[CrossRef][Medline]
- Maeshima A, Niki T, Maeshima A, Yamada T, Kondo H, Matsuno Y. Modified scar grade. A prognostic indicator in small peripheral lung adenocarcinoma. Cancer 2002;95:2546-2554.[CrossRef][Medline]
- Morishita Y, Fukasawa M, Takeuchi M, Inadome Y, Matsuno Y, Noguchi M. Small-sized adenocarcinoma of the lung. Cancer 2001;93:124-131.[Medline]
- World Health Organization Histological typing of lung and pleural tumours. 3rd ed. Berlin: Springer; 1999.
- Hermanek P, Sobin LH. UICC TNM classification of malignant tumors. 4th ed. Berlin: Springer; 1992.
This article has been cited by other articles:
|
|
|
|
 
T. Hashizume, K. Yamada, N. Okamoto, H. Saito, F. Oshita, Y. Kato, H. Ito, H. Nakayama, Y. Kameda, and K. Noda
Prognostic Significance of Thin-Section CT Scan Findings in Small-Sized Lung Adenocarcinoma
Chest,
February 1, 2008;
133(2):
441 - 447.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
N. Yamada, M. Kusumoto, A. Maeshima, K. Suzuki, and Y. Matsuno
Correlation of the Solid Part on High-resolution Computed Tomography with Pathological Scar in Small Lung Adenocarcinomas
Jpn. J. Clin. Oncol.,
December 1, 2007;
37(12):
913 - 917.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Y. Sakao, H. Miyamoto, M. Sakuraba, T. Oh, K. Shiomi, S. Sonobe, and H. Izumi
Prognostic Significance of a Histologic Subtype in Small Adenocarcinoma of the Lung: The Impact of Nonbronchioloalveolar Carcinoma Components
Ann. Thorac. Surg.,
January 1, 2007;
83(1):
209 - 214.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
1. Introduction
