Abstract
Background/Aim: Different tumor markers and systemic inflammation have been linked with cancer development and poor outcome. We aimed to establish a novel non-invasive prognostic index for patients with resectable non-small cell lung cancer (NSCLC) based on serum carcinoembryonic antigen (CEA) and C-reactive protein (CRP). Patients and Methods: Four hundred and sixty-two patients curatively resected for NSCLC between 2008 and 2014 were included. All patients with a follow-up period of less than 5 years were omitted. The geometric mean of the normalized serum CEA and CRP levels was used as a novel tumor marker and inflammation index (TMII). The cut-off value of TMII was determined by receiver operating characteristic (ROC) curve analysis. Univariate and multivariate analyses were used to identify the relative risk factors for survival. Results: ROC curve analysis revealed a TMII cut-off value of 0.46. The group with high TMII displayed more adverse clinical characteristics. Furthermore, compared to patients with low TMII, the group with high TMII had significantly poorer survival. On multivariate analysis, TMII was independently associated with survival. Conclusion: We established a novel prognostic index (TMII) based on serum CEA and CRP. Preoperative TMII may predict poor outcomes in patients with NSCLC.
- Tumor marker
- inflammation index
- carcinoembryonic antigen
- C-reactive protein
- novel index
- non-small cell lung cancer
- surgery
- prognosis
Several studies showed some promising non-invasive prognostic markers of the patients with non-small-cell lung cancer (NSCLC). Serum tumor markers play an important role in cancer diagnosis and prognosis. One such potential marker is carcinoembryonic antigen (CEA) (1, 2). CEA is the most widely used and readily available tumor marker for NSCLC. Previous studies reported statistically significant evidence for the use of serum CEA as a prognostic marker in patients with NSCLC (1, 2).
Furthermore, recent accumulating evidence has shown that systemic inflammation is also linked to poor outcomes in patients with NSCLC (3, 4). Serum C-reactive protein (CRP) is a sensitive indicator of the systemic inflammatory response (5, 6). Previous meta-analyses had shown that an elevated preoperative serum level of CRP was associated with poor prognosis in early-stage NSCLC (5, 6).
Although both CEA and CRP have been reported as prognostic factors, the underlying mechanisms involved are believed to be different. Therefore, we considered that a more useful novel prognostic index might be established using the combination of serum CEA and CRP for operable NSCLC.
Muley et al. introduced an algorithm by which both serum cytokeratin fraction 21-1 (CYFRA 21-1) and CEA can be combined with a new variable for prognostic purposes (7). This algorithm is useful for evaluating not only the degree of marker elevation but also combining the use of two markers. According to this algorithm, in our study, a novel tumor marker and inflammation index (TMII) based on serum CEA and CRP was developed. The prognostic value of TMII in patients with NSCLC who underwent surgery was evaluated.
Patients and Methods
This retrospective study was approved by our Institutional Review Board, and the need to obtain patient consent was waived (IRB Approval No. O-0329). From January 2008 to December 2014 at our hospital, a retrospective study involving 462 patients with resectable NSCLC was conducted. The preoperative serum CEA and CRP levels were also measured at a single laboratory at our hospital. Patients who received preoperative treatment, such as chemotherapy or radiotherapy, those who suffered from any form of inflammatory disease, infection or systemic disease were also excluded. We excluded any patients with NSCLC with a follow-up period of less than 5 years. Data on preoperative laboratory examinations were extracted from our medical records. The median age of the patients was 69 years (range=28-86 years). Among these patients, 250 (54.1%) were males and 212 (45.9%) were females. The overall follow-up periods ranged from 60.7 to 144.7 months.
According to the concept of Muley et al., TMII represents the geometric mean of normalized serum CEA and CRP levels (7). TMII was calculated as follows: √[(serum CEA/5.0 ng/ml) × (serum CRP/0.14 mg/dl)]. A receiver operating characteristic (ROC) curve was plotted to verify the optimum cut-off value for TMII. The area under the curve (AUC) was used as an estimation of diagnostic accuracy.
A comparison between groups with high and low TMII was performed washing the chi-square test. Kaplan–Meier plots were used to depict survival outcomes, with significance determined by the log-rank test. Cox proportional hazards regression model was used to assess the prognostic significance of identified variables in univariable and multivariable analyses. All statistical analyses were performed with the JMP (SAS Institute Inc., Cary, NC, USA). A value of p<0.05 was considered to indicate statistical significance.
Results
Based on the ROC curve analysis, the cut-off value of TMII was defined as 0.46 (Figure 1, AUC=0.709). Patients were subdivided into low and high TMII groups according to this cut-off value. There were 191 patients (41.34%) in the low TMII group; the remaining 271 patients (58.66%) were in the high TMII group. The relationships between TMII and clinicopathological characteristics of the patients are shown in Table I. A positive correlation was found between most of the characteristics examined and a high TMII.
The Kaplan–Meier survival analysis demonstrated that both high serum CEA and CRP were associated with poor cancer-specific survival in patients with NSCLC (5-year survival: normal CEA, 78.3% vs. high CEA: 56.2%, respectively; p<0.001, 5-year survival: low CRP: 79.19% vs. high CRP: 57,14%, respectively; p<0.001) (Figure 2A and B). Similarly, patients with a high TMII had significantly shorter survival as compared to those with low TMII (5-year survival: low TMII: 87.95% vs. high TMII: 61.62%, respectively; p<0.001) (Figure 2C). The difference in survival of groups according to TMII was greater than that according to serum CEA and CRP alone.
In univariate analysis, male gender (p<0.001), current/former smoking status (p<0.001), non-adenocarcinoma histology (p<0.001), pT2-3 status (p<0.001), pN1-2 status (p<0.001), high serum CEA level (p<0.001), high serum CRP level (p<0.001) and high TMII (p<0.001) were associated with poor survival (Table II). All significant prognostic factors tested via multivariate analysis were evaluated using Cox's proportional hazards model. Multivariate analyses regarding postoperative cancer-specific survival in patients with NSCLC were also performed. It was revealed that male gender (p<0.001), non-adenocarcinoma histology (p<0.001), pN1-2 status (p<0.001), and high TMII (p<0.001) were independent predictors of poor prognosis (Table III). Preoperative high TMII remained an independent negative prognostic factor in NSCLC.
Discussion
Serum CEA is the most widely used and readily available tumor marker for NSCLC (1, 2). Previous studies have demonstrated the efficacy of serum CEA as a prognostic marker in patients with NSCLC (1, 2). In the present study, there were significant differences in cancer-specific survival regarding serum CEA level. The mechanism underlying the prognostic role of serum CEA is unknown in detail. However, serum CEA has been reported to be associated with squamous cell carcinoma, moderately to poorly differentiated adenocarcinoma, blood vessel invasion, lymphatic vessel invasion, and visceral pleural invasion (8). All of these features are well known as poor prognostic factors (9-11). Therefore, the prognostic significance of serum CEA may be due to the correlation between serum CEA and the malignant potential of NSCLC.
It is well accepted that there is a strong linkage between inflammation and cancer (3, 4). CRP is a very sensitive indicator of systemic inflammatory response. Published studies have shown that serum CRP is a significant prognostic indicator in NSCLC (5, 6). Patient and tumor factors interact with each other, causing some systemic symptoms, such as pyrexia and cachexia. Their interactions can also accelerate tumor progression or result in tumor regression. A high serum CRP level may reflect a more aggressive tumor type. Thus, the levels of inflammatory components have certain prognostic value in NSCLC, and this theory has been demonstrated by extensive studies (3-6, 12).
Because the mechanisms underlying the prognostic utility of serum CEA and CRP might be different, at least in part, we considered that the combination of preoperative serum CEA and CRP levels might serve as a more accurate and useful prognostic factor. In the present study, we evaluated TMII in patients with NSCLC and demonstrated that TMII is associated with prognosis and can be considered an independent prognostic marker in patients with NSCLC. Furthermore, there was a trend for the difference in survival of groups according to TMII being greater than that according to serum CEA and CRP alone. Thus, TMII was superior to serum CEA or CRP as a predictive factor in patients with NSCLC treated with surgery.
We also showed that high TMII was correlated with male gender, current/former smoking, non-adenocarcinoma subtype, pT2-3 status, pN1-2 status, high serum CEA, and high serum CRP. In other words, the group with a high TMII had more adverse clinical characteristics. Thus, NSCLC with high TMII might have more aggressive tumor behavior. However, the multivariate analysis confirmed the prognostic significance of TMII to be independent of other clinical characteristics.
TMII is easily calculated from results of tests that are routinely available, of low price, and for which results are obtained in a short period of time. Therefore, preoperative TMII might be an inexpensive predictor of survival of patients with NSCLC and should be considered for routine clinical use.
Several limitations should be acknowledged in the current study. Our study is retrospective and has a relatively small sample size from a single center. Further large-scale population-based prospective studies are needed.
Conclusion
In conclusion, we established a novel prognostic index for NSCLC, TMII, based on serum CEA and CRP levels. Preoperative TMII might serve as a useful prognostic marker for patients with NSCLC.
Footnotes
Authors' Contributions
MT conceived the presented idea. MT, RM and TA developed the study design. All Authors contributed to the writing of the final article and approved the article to be published.
Conflicts of Interest
The Authors have no conflicts of interest to declare.
- Received April 30, 2020.
- Revision received May 18, 2020.
- Accepted May 21, 2020.
- Copyright© 2020, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved