Abstract
A proportion of lung cancer patients develop recurrence, even after curative resection. This study reviewed the literature, focusing on the key words: recurrence, relapse, relapse-free survival, disease-free survival (DFS), surgery, and non-small cell lung cancer (NSCLC), not overall survival (OS) to evaluate the post-operative risk factors for tumor recurrence after surgery. This article reviews the current status and progress on this subject.
Lung cancer has been the most common type of cancer world-wide for several decades (1). The standard treatment to cure this disease is surgery. A chest surgeon commonly performs complete resection based on radiographic findings. This type of resection is curative for many patients. However, there are cases that fail to achieve a cure, following surgery. In fact, nearly 50% of patients with non-small cell lung cancer (NSCLC) experience recurrence and have a poor prognosis despite curative resection (2, 3). Therefore, many patients eventually die of their disease (4). TNM staging indicates the level of disease progression and malignant potential of primary lung cancer (5). However, even patients with disease at the same stage are split between the recurrent and non-recurrent group after curative resection. Therefore, the current TNM staging system, which is based on clinical and pathological findings, may have reached the limit of its usefulness (6). Accurately predicting the cases in which disease is, likely to recur can help guide the administration of potentially harmful adjuvant therapies, not only to those most likely to benefit from them (7, 8), but also to those eligible for complete resection, since surgery itself possesses a certain amount of risk (9, 10). There are two methods for the identification of factors related to recurrence following surgery. One is the classical determination using clinical parameters. The other is based of molecular biological techniques. This article reviews the current findings and progress on this subject.
This study searched PubMed, a service of the National Library of Medicine (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi), to identify and extract information published since 2000, regarding factors predictive of recurrence of NSCLC following curative resection. The search focused on the key words recurrence, relapse, relapse-free survival, disease-free survival (DFS), surgery, and NSCLC, not overall survival (OS). The search identified 16 reports of prediction based on clinical parameters (Table I) and 34 based on molecular biological techniques (Table II).
The factors related to recurrence following surgery fall into two categories. One is the classical determination using clinical parameters in addition to the TNM classification. Tumor markers, such CEA were proven to be independent factors for recurrence (11). An extensive pathological investigation is also important. Five Japanese researchers reported that histological differentiation, vessel invasion, lymphatic permeation, and pleural invasion are poor prognostic factors for DFS (11-15). Rena et al. reported that patients with stage I pure bronchial alveolar carcinoma (BAC) have significantly longer DFS than those with similar stage adenocarcinoma (16). On the other hand, three East Asian groups identified the standard uptake value (SUV) as a significant independent factor for DFS (17-19). Furthermore, the physical examination becomes vastly more important because lower performance status (PS) and the presence of symptoms are unfavorable prognostic factors for DFS (20). Williams et al. concurred with these findings (21). Treatment is closely associated with tumor recurrence. Hung et al. reported that treatment for initial recurrence is a significant prognostic indicator in multivariate analyses (22). Surgical procedures other than lobectomy are independently associated with tumor recurrence (23). Another team reported that metastasectomy is a significant favorable factor in patients with distant metastases (24). Furthermore, complete mediastinal lymph node (MLN) dissection is associated with improved survival in comparison to random lymph node sampling for the patients with stage I NSCLC (25). Sugimura et al. reported that pre-operative chemotherapy and post-operative radiotherapy reduced DFS (26). In another study, adjuvant chemotherapy significantly improved DFS (27). Lee et al. used a multiple logistic regression analysis to identify the independent clinical predictors of recurrence and Cox's proportional hazard regression method to successfully develop a clinical prediction model (28). They also developed a predictive system by using an integrated model of clinical information and gene expression (29). The other method for prediction of recurrence uses molecular biology techniques. Lung cancer is a highly aggressive neoplastic disease that includes different histological subtypes with distinct clinicopathological and molecular features (4). Therefore, establishment of useful markers is necessary to accurately classify early- and advanced-stage disease (30). Tumor aggressiveness is related to many genetic alterations associated with cell proliferation, apoptosis, metastasis, and vascularization. An assessment of combination of the expression of cyclin E and p27, the cyclin E-negative/p27-positive group had a significantly higher DFS rate than did other groups (31). Woo et al. also reported that the combined use of KRAS status and the Ki-67 expression level was an excellent prognostic marker to predict the post-operative recurrence of stage I adenocarcinoma (32). The clinical usefulness of Ki-67, but not PCNA in primary lung adenocarcinoma was reported (33). There were also some findings from Western researchers. Poleri et al. showed that MIB-1 and Bcl-2 are independent prognostic factors of recurrence in stage IB but not for stage IA patients (34). Brock et al. reported that methylation of the promoter regions of p16 and CDH13 in both tumor and MLNs are associated with recurrence for patients with stage I NSCLC (35). On the other hand, there have been some negative findings. Baksh et al. reported no statistical significance in a comparison of allelic alterations in genes such as L-myc, hOGG1, APC/MCC, c-fms, p53, and DCC, and point-mutational changes in K-ras-2, with recurrence (36). With relation to metastasis, we previously reported the MACC as a predictive marker for recurrence in primary lung adenocarcinoma by immunohistochemical staining (37) and genetic methods (38). Kinch et al. reported that high levels of EPHA2, which has been linked to the regulation of cellular behavior, in the primary tumor predict brain metastases (39). Higher expression of CXCR7, which plays a role in cancer progression by enhancing tumor growth, is associated with poor DFS in patients with p-stage I NSCLC (40). Strong TS expression is a useful marker for predicting postoperative recurrence following surgery in adenocarcinoma of the lung (41). Several attempts have been made to detect micrometastatic tumor cells or disseminated tumor cells (DTC) in lymph nodes (LNs), bone marrow, and circulating tumor cells (CTCs) in peripheral blood (42,43). CTCs adhere to the endothelial lining and migrate through interstitial spaces in order to metastasize. Sienel et al. reported that MAGE-A transcripts, a marker for disseminated tumor cells, in bone marrow are associated with poor outcome (44). The DFS curves demonstrated that patients with CK cells in the pN0 lymph nodes had significantly shorter survival periods than those without CK cells (45). CK19 mRNA in MLNs is significantly associated with an increased risk of rapid recurrence (46), which is consistent with recent data (4). Qiu et al. also reported that cytoplasmic keratins in lymph nodes are associated with the times-to- relapse (47). However, in contrast to previous published data, Poncelet et al. described that the presence of occult micrometastasis had no influence on DFS (48). Hsu et al.'s study also showed that occurrence of bone marrow microinvolvement is not a good predictor of long-term prognosis (49). D'Amico et al. reported that recurrence is predicted by decreasing levels of E-selectin, increasing levels of CD44, which is a transmembrane glycoprotein identified on cancer stem cells (CSC) as enhancig tumorigenicity (50), and increasing levels of uPAR (51). The importance of tumor-stromal interactions regulating cancer development has been noted. High MMP-2 expression in tumor and stromal cells is associated with poor survival in DFS analyses (52). We previously described a clinical application of dysadherin, a cancer-related cell membrane glycoprotein, and GalNAc-T3, which catalyzes O-glycosylation and has a role in tumor cell binding of cell adhesion molecules (53, 54). Tumor angiogenesis and the expression of angiogenic factors are closely associated with prognosis. The VEGF family of proteins modulates angiogenesis, which is essential for tumor growth and metastasis. However, positive expression of VEGF-C and VEGF were not independent factors associated with recurrence (4). EGFR stimulation activates intracellular signaling cascades that influence angiogenesis. A high co-expression of both IGF1R and EGFR is a significant prognostic factor of poor DFS (55). Activating mutations within the EGFR tyrosine kinase (TK) domain might predict the risk for recurrence in curatively resected pulmonary AD (56). However, two groups noted a lack of prognostic value of EGFR mutations in primary-resected NSCLC (57, 58). Therefore, there is no consensus concerning the role of EGFR mutations in predicting NSCLC recurrence.
IGF is the most important systemic mediator of growth hormone and the IGF pathway has been implicated in NSCLC pathogenesis (59). Shersher et al. reported that low serum IGFBP5 levels strongly correlate with a positive nodal status and any incidence of recurrence (60). There is new information concerning the relationship between IGF/IGF1R and the epithelial-mesenchymal transition (EMT) for the response to IGF1R-inhibitors (61). A significant association was observed between the IGF1R and E-cadherin, γ-catenin, and Ki-67 (62) and, IGF1R predicts postoperative recurrence in patients with adenocarcinoma (59). EMT is an increasingly recognized mechanism to generate further CSC endowed with a more invasive phenotype (63, 64). In fact, the combination of stem cell markers CD133 and ABCG2 is reported to predict a relapse in stage I NSCLC (65).
Clinical parameters predicting recurrence after complete resection for patients with NSCLC as seen in current literature.
Cancer is thought to arise from the multistage accumulation of not only genetic but also epigenetic alterations. Therefore, there are various techniques to investigate epigenetic changes to improve risk stratification. ALDH2, TPM3, ECH1, and IMMT were found by proteomic screening being able to predict tumor recurrence (66). Microarray analysis has made it possible to simultaneously measure the expression of thousands of genes (6). Tumor-specific genetic fingerprints/gene signatures can affect the prognosis (67). The deregulation of micro RNAs is linked to cancer initiation and progression, indicating that miRNAs may act as tumor suppressor genes or oncogenes (7). In fact, micro RNA expression in resected NSCLC could potentially identify those at high-risk of relapse after surgery (68). Single-nucleotide polymorphisms of DNA- and histone-modifying genes might be used as predictive biomarkers not only to identify patients with stage I NSCLC, who could benefit from adjuvant chemotherapy, but also to predict response to adjuvant chemotherapy (69). Despite these findings, it still remains a challenge to assign a prognostic value to these molecular pathways because of poor reproducibility of findings (70). Aberrant methylation of CpG islands acquired in tumor cells in promoter regions can cause the loss of gene function. Demethylation of DAPK from normal tissues and hypermethylation of RARbetaP2 from normal tissues are risk factors for a reduced DFS (71).
Molecular parameters to predict recurrence.
There are three limitations for the interpretation of these findings: [i] There is uncertainty differentiating between a second primary lung cancer and recurrent primary tumors. The molecular approaches might powerful methods to resolve this problem (72,73). [ii] Recurrent cases might include slow growth not leading to death, and non-recurrent cases may include the small lesions that cannot be detected (74). [iii] There is abundant heterogeneity in the findings in this review because most studies included various stages of NSCLC. Furthermore, individual studies are typically statistically underpowered with regard to their conclusion (5). Therefore, clinical trials with recurrence as a primary endpoint are needed in order to overcome these limitations.
Acknowledgements
This study was supported, in part, by a Grant-in-Aid for Scientific Research (C) 21591831 from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan and a UOEH Research Grant for Promotion of Occupational Health.
Footnotes
-
Conflict of Interest
The Authors declare no conflict of interest.
- Received May 2, 2012.
- Revision received July 16, 2012.
- Accepted July 17, 2012.
- Copyright© 2012 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
