Ann Thorac Surg 2004;78:1004-1009
© 2004 The Society of Thoracic Surgeons
Original article: general thoracic
Effect of histologic type and smoking status on interpretation of serum carcinoembryonic antigen value in non–small cell lung carcinoma
Morihito Okada, MD, PhDa,*,
Wataru Nishio, MD, PhDa,
Toshihiko Sakamoto, MD, PhDa,
Kazuya Uchino, MDa,
Tsuyoshi Yuki, MDa,
Akio Nakagawa, MDa,
Noriaki Tsubota, MD, PhDa
a Department of Thoracic Surgery, Hyogo Medical Center for Adults, Akashi City, Hyogo, Japan
Accepted for publication March 8, 2004.
* Address reprint requests to Dr Okada, Department of Thoracic Surgery, Hyogo Medical Center for Adults, Kitaohji-cho13-70, Akashi City, 673-8558, Hyogo, Japan
morihito1217jp{at}aol.com
 |
Abstract
|
|---|
BACKGROUND: Serum carcinoembryonic antigen (CEA) has all of the properties desired for a biologic measure to be used as a prognostic indicator in the clinical evaluation of lung cancer. Carcinoembryonic antigen value appears to be related to tumor histologic type and patients' smoking status, which has yet to be intensively analyzed as reports available thus far have consisted of a limited number of patients. This study was undertaken to determine whether the prognostic value of CEA differs according to histologic type in a large group of patients with clinical early-stage lung cancer, and how smoking influences its value.
METHODS: Two series of 694 and 260 consecutive patients who underwent resection for clinical stage I lung adenocarcinoma and squamous cell carcinoma, respectively, were evaluated. We measured serum CEA before and after surgery, and analyzed its prognostic significance in relation to histologic type and its correlation with smoking status.
RESULTS: We found significantly higher CEA levels in patients with adenocarcinomas than in those with squamous cell carcinomas (7.8 versus 5.5 ng/mL; p = 0.0018), but a higher percentage of CEA-positive patients among those with squamous cell carcinoma (109 of 260, 41.9%) than those with adenocarcinoma (245 of 694, 35.3%). Clinical stage I patients with a high preoperative CEA level had a poor prognosis, and for pathologically confirmed stage I patients with a high postoperative CEA level the prognosis was worse. The prognostic value of serum CEA level was thus significantly greater for adenocarcinoma than for squamous cell carcinoma. This was probably because of a much higher proportion of smokers among patients with squamous cell carcinoma. In adenocarcinoma, the growth of which was generally less influenced by smoking, the proportion of CEA-positive smokers (49.3%, 170 of 345) was greater than that of CEA-positive nonsmokers (21.5%, 75 of 349, p < 0.0001). Additionally, in patients with adenocarcinoma, survival of nonsmokers was more greatly influenced by CEA level than that of smokers.
CONCLUSIONS: Although serum CEA values measured before and after surgery are important in identifying patients at high risk of poor survival, its specificity is higher for adenocarcinoma than for squamous cell carcinoma. When serum CEA levels are checked, smoking status of patients, particularly of those with squamous cell carcinoma, should be taken into account.
|
Introduction
|
|---|
In 1997, the American Thoracic Society and the European Respiratory Society jointly stated the guidelines for evaluation of non–small cell lung cancer [1]. It was noted that no serum tumor marker was sensitive and specific enough to identify occult disease or influence treatment, and that routine measurement of biomarkers was not recommended for screening or staging of disease. Thus, the role of serum carcinoembryonic antigen (CEA) levels measured before or after resection of non–small cell lung cancer is not widely accepted. However, several studies have indicated that CEA in patients with non–small cell lung cancer provides information useful for determining survival independent of the stage of the disease [2–7].
Non–small cell lung cancer constitutes a histologically heterogeneous group of lung cancers, among which the two major subtypes are adenocarcinoma and squamous cell carcinoma. The majority of studies done thus far on the prognostic value of serum CEA involved a patient population not stratified by histologic type. There are few reports examining the differences in CEA prognostic values between adenocarcinoma and squamous cell carcinoma. Occasionally serum CEA levels are elevated in patients with nonmalignant diseases such as chronic bronchitis, emphysema, or colitis [8, 9]. In addition, cigarette smoking is one of the most powerful variables associated with increased serum CEA levels [10, 11]. We undertook this study to evaluate the prognostic value of CEA levels in patients with adenocarcinoma and squamous cell carcinoma, as well as to investigate how the significance of serum CEA is affected by smoking.
|
Material and methods
|
|---|
From January 1985 through December 2002, two series of 694 and 260 consecutive patients with clinical stage I disease were operated on for proven primary adenocarcinoma and squamous cell carcinoma of the lung, respectively. The histologic type of tumor was determined by applying the World Health Organization classification. In all patients, we measured serum CEA before and after surgery and resected the primary tumor. The tumor was measured directly in the surgical specimens. For preoperative clinical staging, we used a detailed history and physical examination, biochemical profile, chest roentgenogram, bronchoscopy, computed tomography of the chest, brain, and upper portion of the abdomen, and bone scintigraphy. Stage was determined according to the international TNM staging system [12]. Patients who had undergone preoperative chemotherapy or radiotherapy were excluded.
All thoracotomies were performed within 1 month of the preoperative CEA measurement, and postoperative CEA was measured 1 month after surgery. The serum CEA was determined by an enzyme immunoassay (Fuji Rebio, Tokyo, Japan). According to the manufacturer, the average for this assay in healthy individuals is 1.93 ng/mL and the upper limit of normal is 5.0 ng/mL. Generally, the patients were postoperatively examined at 3-month intervals for 5 years and thereafter at 1-year intervals to check for recurrence and survival. We defined smokers as patients who were smoking at the time of diagnosis of lung cancer. The nonsmokers group thus included ex-smokers as well as patients who had never smoked.
The statistical significance of differences among the subdivided groups and several clinicopathologic variables was analyzed by Mann-Whitney U test. Survival was calculated by the Kaplan-Meier method, and differences in survival were determined by log-rank analysis. A multivariable analysis of several prognostic factors was carried out using Cox's proportional hazards regression model. Zero time was the date of pulmonary resection, and the terminal event was death attributable to cancer, noncancer, or unknown causes. Significance was defined as p less than 0.05.
|
Results
|
|---|
Clinical characteristics, surgical treatment, and pathologic stage are summarized in Table 1. Compared with squamous cell carcinoma patients, adenocarcinoma patients were younger (p < 0.0001), included a higher number of women (p < 0.0001), had a smaller tumor (p < 0.0001), and had higher levels of preoperative serum CEA (p = 0.0018). However, the percentage of preoperative CEA-positive patients was higher in the squamous cell carcinoma group (41.9% versus 35.3%). The two groups were fairly similar with respect to surgical procedure, resectability of tumor, and pathologic stage.
Overall follow-up ranged from 9 to 225 months, with a median of 62 months for surviving patients. The 5-year survival rates for adenocarcinoma versus squamous cell carcinoma were 66.8% versus 66.2% in patients with clinical stage I disease (Fig 1A). There were no significant differences between the two groups (p = 0.5721). In contrast, in patients with pathologic stage I disease, the 5-year survival rates for adenocarcinoma versus squamous cell carcinoma were 81.1% versus 70.3%, respectively (Fig 1B). This difference was significant (p = 0.0075). When the disease was diagnosed as proven stage I, patients with adenocarcinoma had better survival than those with squamous cell carcinoma. These data suggested that adenocarcinoma judged preoperatively as clinical stage I disease included more advanced disease compared with squamous cell carcinoma, and that preoperative staging for adenocarcinoma was consequently more difficult.
View larger version (23K):
[in this window]
[in a new window]
|
Fig 1. Overall survival curves for patients with clinical (A) and pathologic (B) stage I non–small cell lung cancer distributed according to histologic type.
|
|
We then analyzed the effect of preoperative serum CEA level on survival of clinical stage I patients. Among patients with low CEA level (
2.5 ng/mL), the survival rate was significantly better for those with adenocarcinoma than for those with squamous cell carcinoma (p = 0.0417). Among patients with a high CEA level (>5.0 ng/mL), the survival rate of those with adenocarcinoma appeared poorer (not significant, p = 0.1218), although the survival rates of the two groups were similar (p = 0.6457) for those with intermediate CEA levels (2.6 to 5.0 ng/mL). For adenocarcinoma patients, high preoperative CEA levels were associated with a significantly poorer survival (p < 0.0001; Fig 2A), the statistical difference of which was definitely reduced in squamous cell carcinoma patients (p = 0.1166; Fig 2B). These data indicated the prognostic value of serum CEA before surgery was even more important for adenocarcinoma than for squamous cell carcinoma.
View larger version (21K):
[in this window]
[in a new window]
|
Fig 2. Overall survival curves for patients with clinical stage I adenocarcinoma (A) and squamous cell carcinoma (B) distributed according to preoperative serum carcinoembryonic antigen (CEA) level.
|
|
Univariate analyses to identify important preoperative variables for prognosis demonstrated that male sex (p < 0.0001), age older than 65 years (p < 0.0001), tumor size larger than 30 mm (p < 0.0001), current smoker (p = 0.0022), and high CEA level (p < 0.0001), but not histologic type (p = 0.5475), significantly and negatively affected survival of patients with clinical stage I disease. To better evaluate these factors affecting survival, multivariate analysis was performed (Table 2). Male sex (p = 0.0092), older age (p = 0.0001), and larger size of tumor (p < 0.0001) strongly predicted failure. Serum CEA status was also an independent significant factor predictive of the outcome (p = 0.0401). The multivariate test demonstrated that patients with adenocarcinoma had a marginally (not significant) worse prognosis than those with squamous cell carcinoma (p = 0.0515), and that there was no difference in survival between smokers and nonsmokers (p = 0.07829).
Next, we examined the effect of postoperative serum CEA level on survival of patients with pathologically confirmed stage I disease. Among patients with either adenocarcinoma or squamous cell carcinoma, those with a high CEA level had a significantly worse prognosis (p = 0.0003 and p = 0.0016, respectively; Fig 3). The 5-year survival rates of patients with adenocarcinoma and squamous cell carcinoma and a high CEA level were 57.4% and 28.0%, respectively. These data suggested the importance of postoperative serum CEA value even when pathologic examination revealed stage I disease.
View larger version (21K):
[in this window]
[in a new window]
|
Fig 3. Overall survival curves for patients with pathologic stage I adenocarcinoma (A) and squamous cell carcinoma (B) distributed according to postoperative serum carcinoembryonic antigen (CEA) level.
|
|
To detect the effect of cigarette smoking at the time of diagnosis on CEA tests, the correlation between smoking status and serum CEA values was investigated (Table 3). In patients with adenocarcinoma, smoking is generally believed to play a less vital role in the cause or growth of the tumor compared with other types. Of 694 patients with adenocarcinoma, 345 were smokers (49.7%) and 349 were nonsmokers (50.3%). The ratio of CEA-positive smokers (49.3%, 170 of 345) was significantly higher than that of CEA-positive nonsmokers (21.5%, 75 of 349; p < 0.0001), suggesting that serum CEA level was influenced at least in part by smoking. In addition, 92.3% (240 of 260) of patients with squamous cell carcinoma were smokers, implying that the clinical significance of serum CEA values should be carefully determined, especially in patients with squamous cell carcinoma. Finally, to investigate the influence of smoking on CEA level, we analyzed the survival of smokers and nonsmokers with clinical stage I adenocarcinoma distributed according to their preoperative CEA level (Fig 4). Although nonsmokers with a high CEA value had a significantly worse prognosis than those with a normal CEA value (p < 0.0001), this significant difference disappeared in smokers (p = 0.1264). These data suggested the prognosis of nonsmokers with adenocarcinoma was more significantly affected by the CEA level than that of smokers with adenocarcinoma.
View this table:
[in this window]
[in a new window]
|
Table 3. Relationship Between Preoperative Carcinoembryonic Antigen and Smoking Status in Patients With Clinical Stage I Disease
|
|
View larger version (20K):
[in this window]
[in a new window]
|
Fig 4. Overall survival curves for nonsmokers (former/never) (A) and current smokers (B) with clinical stage I adenocarcinoma distributed according to preoperative serum carcinoembryonic antigen (CEA) level.
|
|
|
Comment
|
|---|
Preoperative serum CEA value had independent prognostic value after adjusting for sex, age, tumor size, and histologic type as evaluated in patients treated surgically for clinical stage I non–small cell lung cancer. Its increase was related to clinically and statistically significant reduction in survival even after intentional curative resection for early-stage disease. Although serum CEA value does not appear to be a specific marker of lung cancer, it is an essential prognostic factor. Thus, our data obtained from a large number of patients verified the importance of preoperative CEA as an adjuvant factor to conventional ones used for preoperative clinical staging despite modern progress in imaging and diagnostic procedures.
The relationship between serum CEA values and tumor histologic type remains obscure. The most interesting issue examined in this study was whether a histologic difference between adenocarcinoma and squamous cell carcinoma could affect serum CEA value. Interestingly, although serum CEA levels were significantly higher in patients with adenocarcinoma than in those with squamous cell carcinoma, the proportion of CEA-positive patients with adenocarcinoma (35.3%, 245 of 694) was less than that of CEA-positive patients with squamous cell carcinoma (41.9%, 109 of 260). Based on these data, we speculated that the majority of CEA-positive patients with squamous cell carcinoma had marginally positive levels of CEA, and consequently that the specificity of the CEA test was low for squamous cell carcinoma patients. In addition, based on survival analyses (Figs 2, 3), we considered that the CEA test was more accurate and more specific for adenocarcinoma than for squamous cell carcinoma. The difference in specificity of CEA testing between adenocarcinoma and squamous cell carcinoma was thought to depend basically on the biologic nature of each tumor. However, we could not rule out the possibility that other factors might contribute to this difference.
Because the factor most strongly influencing the increase of serum CEA was reported to be cigarette smoking [10, 11], we quantified the effect of smoking on serum CEA level. Among nonsmokers, the rate of CEA-positive patients was 21.5% (75 of 349) for adenocarcinoma and 20.0% (4 of 20) for squamous cell carcinoma. In contrast, among smokers, the rate of CEA-positive patients was 49.3% (170 of 345) for adenocarcinoma and 43.8% (105 of 240) for squamous cell carcinoma. These data suggested that in more than half of CEA-positive smokers, serum CEA was increased by cigarette smoking. On the other hand, although the proportion of smokers among adenocarcinoma patients was 49.7% (345 of 694), that among squamous cell carcinoma patients was 92.3% (240 of 260). That is why the specificity of serum CEA tests was low for squamous cell carcinoma. Besides, the survival analysis showed that the CEA level more significantly influenced the outcome of nonsmokers with adenocarcinoma than that of smokers with adenocarcinoma, confirming that CEA was not necessarily valid in current smokers. These data suggested that the value of CEA as a marker should be considered in relation to the histology of non–small cell lung cancer subtypes and smoking status.
Postoperative serum CEA values, the role of which did not appear to be influenced by the histologic type, provided valuable information regarding patient prognosis after surgery. Even though our analysis was limited to pathologically proven stage I disease, the increase in postoperative serum CEA was associated with poor survival rates, which for 5 years were 57.4% and 28.0% for adenocarcinoma and squamous cell carcinoma, respectively. These data indicated that measurement of postoperative serum CEA, as well as preoperative CEA, provides information useful for detecting patients at high risk of poor survival; consequently in patients with high serum CEA levels, additional adjuvant therapy might have to be considered.
|
References
|
|---|
- The American Thoracic Society and The European Respiratory Society. Pretreatment evaluation of non–small-cell lung cancer. Am J Respir Crit Care Med. 1997;156:320–322[Free Full Text]
- Concannon JP, Dalbow MH, Hodgson SE, et al. Prognostic value of preoperative carcinoembryonic antigen (CEA) plasma levels in patients with bronchogenic carcinoma. Cancer. 1978;42:1477–1483[Medline]
- Ford CH, Stokes HJ, Newman CE. Carcinoembryonic antigen and prognosis after radical surgery for lung cancer: immunocytochemical localization and serum levels. Br J Cancer. 1981;44:145–153[Medline]
- Dent PB, McCulloch PB, Wesley-James O, MacLaren R, Muirhead W, Dunnett CW. Measurement of carcinoembryonic antigen in patients with bronchogenic carcinoma. Cancer. 1978;42:1484–1491[Medline]
- Sawabata N, Ohta M, Takeda S, et al. Serum carcinoembryonic antigen level in surgically resected clinical stage I patients with non–small cell lung cancer. Ann Thorac Surg. 2003;74:174–179
- Okada M, Sakamoto T, Nishio W, Uchino K, Tsubota N. Characteristics and prognosis of patients after resection of nonsmall cell lung carcinoma measuring 2 cm or less in greatest dimension. Cancer. 2003;98:535–541[Medline]
- Okada M, Nishio W, Sakamoto T, et al. Prognostic significance of perioperative serum carcinoembryonic antigen in non–small cell lung cancer. Ann Thorac Surg. In press
- Fujishima T, Honda Y, Shijubo N, Takahashi H, Abe S. Increased carcinoembryonic antigen concentrations in sera and bronchoalveolar lavage fluids of patients with pulmonary alveolar proteinosis. Respiration. 1995;62:317–321[Medline]
- Rule AH, Straus E, Vandevoorde J, Janowitz HD. Tumor-associated (CEA-reacting) antigen in patients with inflammatory bowel disease. N Engl J Med. 1972;287:24–26
- Stevens DP, Mackay IR. Increased carcinoembryonic antigen in heavy cigarette smokers. Lancet. 1973;2:1238–1239[Medline]
- Alexander JC, Silverman NA, Chretien PB. Effect of age and cigarette smoking on carcinoembryonic antigen levels. JAMA. 1976;235:1975–1979[Abstract]
- Mountain CF. Revisions in the International System for Staging Lung Cancer. Chest. 1997;111:1710–1717[Abstract/Free Full Text]
Related Article
-
Invited commentary
- Marc de Perrot
Ann. Thorac. Surg. 2004 78: 1009-1010.
[Extract]
[Full Text]
[PDF]
This article has been cited by other articles:
|
|
|
|
 
H. Matsuguma, R. Nakahara, S. Igarashi, Y. Ishikawa, H. Suzuki, N. Miyazawa, S. Honjo, and K. Yokoi
Pathologic stage I non small cell lung cancer with high levels of preoperative serum carcinoembryonic antigen: Clinicopathologic characteristics and prognosis
J. Thorac. Cardiovasc. Surg.,
January 1, 2008;
135(1):
44 - 49.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. Okada, S. Tauchi, K. Iwanaga, T. Mimura, Y. Kitamura, H. Watanabe, S. Adachi, T. Sakuma, and C. Ohbayashi
Associations among bronchioloalveolar carcinoma components, positron emission tomographic and computed tomographic findings, and malignant behavior in small lung adenocarcinomas
J. Thorac. Cardiovasc. Surg.,
June 1, 2007;
133(6):
1448 - 1454.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
