DOI: https://doi.org/10.21203/rs.2.15414/v1
Background The prognostic significance of cyclooxygenase-2 ( Cox-2 ) expression has been implicated in various cancers such as carcinomas of the breast, head and neck squamous cell, oral squamous cell, and nasopharyngeal. Despite meta-analysis studies of Cox-2 expression in Head and Neck cancer, similar investigation that focus on nasopharyngeal carcinoma (NPC) remains unexplored. Thus far, expression studies of Cox-2 in NPC have not yield conclusive findings due to conflicting results. This study, therefore, is the first to evaluate Cox-2 expression with the survival outcome and treatment response of NPC patients via a systematic meta-analysis approach.Methods Meta-analysis was done in compliance with the standard protocol of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Checklist. Potential studies were obtained from PubMed, Science Direct and Scopus. Fundamental and primary clinical characteristics of patients and hazard ratios (HRs) with 95% confidence intervals (CIs) of overall survival (OS) data were then collected and tabulated. Analysis of relationship between Cox-2 expression and survival outcome in NPC patients were explained with the aid of forest plot charts. Odd ratio (OR) evaluation to determine the relationship between Cox-2 expression and treatment response was via MedCalc (version 18.11) software, with results displayed in forest plots.Results Seven eligible studies containing 495 NPC patients were finally used in the meta-analysis. The pooled HRs with 95% confidence intervals (CIs) for OS is 2.53 [2.13, 3.27], indicating that overexpression of Cox-2 corresponds to poor survival. In the case of treatment response, the pooled OR is 0.98 [0.27-3.49], revealing that the expression pattern of Cox-2 did not relate to the outcome of treatment.Conclusions Overexpression of Cox-2 predicts worse survival outcome for NPC patients, but not treatment response. Hence, the prognostic value of Cox-2 in NPC is only relevant when survival probability of patients is considered.
Nasopharyngeal carcinoma (NPC) is a form of malignancy at the tissue of the upper section of pharynx behind the nose [1] – the nasopharynx region. Globally, it accounts for 65,000 deaths yearly but also has a regionally varied incidence rate [2]. In endemic regions such as Southern China, Southeast Asia and the Middle East, there are over 20 cases per 100,000 people albeit rare in North America and Europe [3]. Based on World Health Organization (WHO) classification, NPC is histologically categorized as three subtypes, namely Type I: keratinizing differentiated squamous cell carcinoma, Type II (or 2a): non-keratinizing differentiated squamous cell carcinoma, and Type III (or 2b): non-differentiated undifferentiated basaloid squamous cell carcinoma [4]. The more chemo-sensitive type (Type III/2b) is predominant among Asian cases, whereas Types I and II/2a are mostly found in Western countries [5].
Risk factors for NPC include genetic, viral (Epstein–Barr Virus, EBV) infection, environment, lifestyle (smoking), and consumption of certain preserved food [6]. Early stages of malignancy usually involve invasion of NPC cells to surrounding tissue and cervical lymph nodes [7, 8]. Despite the radio-sensitivity of NPC tumors, patients with advanced stages show poor survival rate [9 – 12]. Improved therapeutic techniques such as concurrent chemotherapy with or without neo-adjuvant/adjuvant, intensity-modulated radiation therapy (IMRT), and high resolution magnetic resonance image (MRI) monitoring have been the standard treatment protocol for locally advanced NPC [13 - 15]. Nevertheless, relapse and metastasis still occur in approximately 20-50% of patients [16].
Studies have shown that NPC patients in the same classification present with different prognosis [17, 18]. This suggests that ethnicity consideration [19] and bio-molecular factors associated with survival outcome [20] may be necessary to accurately distinguish NPC patients for individualized and tailored treatment that is more effective. As such, identifying prognostic factors, particular molecular/genetic factors, that correspond closely to the actual clinical outcomes is necessary for improvement of therapies that yield better treatment outcome.
One such candidate biomarker of potential prognostic significance to NPC is the Cyclooxygenase-2 (Cox-2) gene. Its expression has clinical and prognostic significance in cancers of the head and neck [21]. More specifically in NPC, its involvement in carcinogenesis is most probably via participation in the formation of inflammatory microenvironment that has consequence on tumorigenesis and malignancy [22]. Cox-2 overexpression is also associated with metastasis to lymph node in NPC patients [23]. Its most probably mediates metastasis in NPC by promoting interactions between cancer and myeloid-derived suppressor cells [24].
Cox-2, one of the two isoforms of Cox gene and a key enzyme in the conversion of arachidonic acid to prostaglandins, is known to be poorly expressed (if not detectable at all) in most normal tissues but is rapidly induced by pro-inflammatory cytokines, growth factors, carcinogens and tumor promoters [25 - 28]. Its overexpression in various cancers has been linked to angiogenesis, invasion, and apoptotic resistance - suggesting its involvement in inflammation-induced tumorigenesis, and its influence on the outcome and survival of cancer cells [24, 26, 29 - 33]. Interestingly, the functional abrogation of Cox-2 via Cox inhibitor reverses cancer progression [34 - 37]. Actions of Cox inhibitor and non-steroidal anti-inflammatory drugs (NSAIDs) have effects on the reduced incidence and progression of tumors in animal models [38].
To date, a comprehensive review analysis of Cox-2 as per its prognostic value for NPC is lacking. Inferences from existing literature linking its expression to NPC tissues are compounded by the fact that the experimental studies differ in methodology, and each of them was biased by relatively small sample size. As a result, findings have not been consistent. Herein, we report a systematic review and meta-analysis study to established the relationship between Cox-2 expression and survival outcome and treatment response in NPC patients.
Literature Search
Exhaustive literature search was done via electronic and manual strategies. In the former, searches were conducted on PubMed (https://www.ncbi.nlm.nih.gov/pubmed/), Science Direct (https://www.sciencedirect.com/) and Scopus (https://www.scopus.com/). For the latter, manual searches were done using the reference lists of relevant articles. The keywords used to retrieve related articles and abstracts were Nasopharyngeal Carcinoma (NPC) and Cyclooxygenase-2 (Cox-2). The currency of collected literature data was as at March 7, 2019.
Selection Criteria
Titles and abstracts of all the articles retrieved from literature search strategy were scrutinized for relevancy based on specific inclusion and exclusion criteria. The inclusion criteria comprised studies (1) containing NPC patients, (2) evaluating the expression of Cox-2, (3) with data related to hazard ratio (HR) and/or odd ratio (OR) at 95% confidence interval (CI), (4) reported in English, and (5) done in human. Exclusion criteria consisted of studies (1) from review articles or letters, (2) written in Chinese, (3) with duplicated data, (4) containing insufficient information to calculate the log hazard ratio (logHR), standard error of logHR (SE) and/or odd ratio (OR) for analysis, (5) from meta-analysis articles, and (6) of single nucleotide polymorphism (SNP) related reports. Following selection, full text articles of potential eligible studies were downloaded to procure the required data. To avoid duplication, only the newest article with biggest samples size was included when more than a trial was done within the same patient population.
Data retrieval
Data extracted from each eligible studies included first author’s last name, publication year, region of publication, ethnicity, number of female patients, number of patients with high/low Cox-2 expression, Cox-2 expression assay (method and cut-off level), clinic-pathological data (number of patients with different tumor stages), therapy regimen used, and survival data (HR at 95% CI for OS and OR).
Statistical Analysis
Primary results of this meta-analysis were overall survival (OS) rate, which was expressed as hazard ratio (HR) at 95% confidence intervals (CIs). The OS is a direct estimate of clinical benefit to a patient, and is defined as the time from random assignment to death [39]. LogHR and SE data available in any eligible studies were used directly for the purpose of our meta-analysis. In instances where the logHR and SE were not shown directly, methods developed by Parmar et al. [40] was used to obtain the relative data. Relevant methods provided in the same studies were also used to determine the pooled HR, CI and weight. Basically, HR of more than one (HR > 1), reflects a poor prognosis in Cox-2 over-expression. Significance of the outcome was proven at p value <0.05. Significant heterogeneity exists in pooled HRs when p < 0.10 or I2 > 50%. Results for each individual study and the pooled results for included eligible studies were displayed as forest plot. The OR was to determine the impact of Cox-2 expression on treatment response. MedCalc (version 18.11) software was used to determine the OR at 95% CI. Forest plot of OR for included studies was also graphed. The same software was also used to determine the publication bias, of which was presented by I2 and p-value. Significant heterogeneity was taken for pooled ORs of p < 0.10 or I2 > 50%, and graphically represented using a funnel plot. Publication bias is taken for p > 0.10, and can be visually evaluated by the asymmetry of the funnel plot.
Eligible Studies
In the initial search, 46, 24 and 37 potential studies were obtained from Scopus, Science Direct and PubMed respectively, yielding a total of 107 studies (Figure 1). From these, 53 studies were excluded after their titles and abstracts were reviewed and deemed irrelevant. Full texts of the remaining 54 potentially relevant studies were downloaded and reviewed in detail. Subsequently, another 36 studies were excluded for the reason that nine were written in Chinese and 27 were duplicates. Eleven of the remaining 18 articles have incomplete data for meta-analysis, leaving a final of 7 eligible studies (Figure 1).
The publication years for these 7 eligible studies are between 2009 and 2018, and include a total of 495 patients (range of 38 to 148 per study). Retrieved clinical information from these patients is shown in Table 1. All the studies involved cases in Asia, and hence only among Asian patients. Except for one study, female patients represent a minority group among the total number of patients admitted. In 6 of the studies, a majority of the patients presents with advanced stages of cancer (Stage III & IV) upon diagnosis. Albeit a variety of chemotherapy strategies, the dominant treatment method reported in 6 of the studies is radiotherapy. With the exception of one study, the detection method for Cox-2 expression was Immunohistochemistry assay. Four of the studies have more patients in the high expression group. All 7 articles provided data (OS) for survival analysis, whereas only two of the 7 articles have data (OR) for analysis of treatment response.
Correlation between Cox-2 expression and survival outcome
Correlation of NPC patients’ survival outcome with Cox-2 expression was determined by the hazard ratio (HR) value of OS. The pooled HR (95% CI) of OS from the 7 studies is 2.63 [2.13, 3.27], (I2 = 93%, P < 0.001) in a random model (Figure 2). Patient group with higher Cox-2 expression shows higher hazard of death compared to the control groups as the HR obtained was greater than one. The HR value of 2.63 implies that at any time during the follow-up, patient group with Cox-2 over-expression have a 163% higher risk of death. OS was statistically significant because the 95% CI did not include 1. However, heterogeneity of our result is significant, suggesting that our findings may be improved further by increasing sample size.
Correlation between Cox-2 expression and treatment response
Two of the 7 studies that has sufficient data on treatment response were analysis to determine correlation between Cox-2 expression and treatment response in NPC patients. Odd ratio (OR) results were evaluated on this case. The pooled OR (at 95% CI) was 0.98 [0.27, 3.49], (I2 = 0%, P = 0.99) in a random model (Figure 3). Since the OR is less than 1, the treatment response among patients with high Cox-2 expression was better than the control group. Nonetheless, this result is not statistically significant because the 95% CI included 1.
Assessment of publication biasness
Based on the funnel plot (OR vs standard error, Figure 4) no publication bias is evident. Hence, our result is significant, and closely relevant to true scenario.
The main finding of our meta-analysis is the fact that over-expression of Cox-2 is associated with poor prognosis of NPC patients with respect to survival outcome. In fact, patients with high Cox-2 level are 2.63 times more likely to die post-treatment relative to those in the low expression group. This observation of correlation between poor prognosis and up-regulation of Cox-2 has also been reported in other previous studies of different cancers. Its over-expression has been linked to angiogenesis and lymph node metastasis in breast carcinoma [41], tumor progression in head and neck squamous cell carcinoma (HNSCC) [42], and oral squamous cell carcinoma (OSCC) [43]. There has only been one other study that contradicts this observation. Loong et al. [44] reported that weak or low (rather than high) Cox-2 expression was associated with worse survival rate in NPC. Nevertheless, the nature of our analysis that involved pooled data from multiple studies on NPC validates our conclusion, and affirms the reliability of Cox-2 as a prognostic marker for NPC.
In the event of NPC carcinogenesis, up-regulation of Cox-2 most probably play essential role(s) in angiogenesis, invasion, metastasis and inhibition of apoptosis – demanding its persistence presence in a wide range of pre-neoplastic and malignant conditions [34-38]. Its induction by inflammatory cytokines, tumor promoters, and growth factors in cancer cells/tissues [45] helps to sustain its level in cancer progression where it acts in concert with VEGF and EGRF, that are similarly activated [46, 47]. Incidentally, the tumorigenic effects of Cox-2 are inhibited by selected Cox-2 inhibitors [24, 41, 48]. One of such inhibitors is Celecoxib which demonstrated anti-proliferative, anti-invasive and anti-angiogenic effects on an NPC cell line in a dose-dependent manner [49]. Potentially, Celecoxib could be a use in combination with current chemo- and radiotherapy strategies in the treatment of NPC patients – a prospect that warrants clinical trial studies.
For treatment response, our analysis reveals that patients with high Cox-2 expression responded better compared to low expression or control group. However, our result is statistically insignificant, hence inconclusive. A contributing factor to this problem could be the very limited number of studies (only two) available for meta-analysis. It remains to be explored whether chemo- and/or radiotherapy responsiveness among NPC patients with high Cox-2 level is a prevalent observation. More clinical and experimental studies on this matter will be needed to resolve this question.
Quintessentially, our study reveals the correlation of low survival rate among NPC patients with high Cox-2 expression. This is the first study that scientifically established the relevance of Cox-2 as a prognostic biomarker for NPC as far as post-treatment survival outcome is concerned. Hence, development of a standard protocol for the evaluation of Cox-2 expression in NPC patients following chemo- and radiotherapy should now have direct applicability in the management and treatment nasopharyngeal cancer.
Survival rate of NPC patients correlates negatively with expression level of Cox-2. However, there is no significant correlation between Cox-2 expression and treatment outcome. Therefore, Cox-2 can only be a prognostic biomarker for NPC in the assessment of survival probability.
5-FU: 5-fluorouracil
ASR: age-adjusted rate
CCRT: Concurrent chemoradiotherapy
CIs: confidence intervals
CRT: chemoradiotherapy
Cox-2: cyclooxygenase-2
EBV: Epstein–Barr virus
HRs: hazard ratios
HNC: Head and Neck cancer
HNSCC: head and neck squamous cell carcinoma
IC: induction chemotherapy
ICRT: induction chemoradiotherapy
IHC: Immunohistochemistry
IMRT: intensity-modulated radiation therapy
ISH: RNA-in situ hybridization
I2: I-squared
MRI: magnetic resonance imaging
NPC: Nasopharyngeal Carcinoma
NSAIDs: non-steroidal anti-inflammatory drugs
NR: not related
ORs: odd ratios
OSCC: oral squamous cell carcinoma
OS: overall survival
PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses
RT: radiotherapy
SE: standard error
WHO: World Health Organization
Ethics approval and consent to participate
Not applicable
Consent for publication
Not applicable
Availability of data and materials
The datasets used (including data generated) and/or analyzed during this study are included in this published article.
Competing interests
The authors declare that they have no competing interests.
Funding
This study is funded by the Malaysian Ministry of Education via the Trans-disciplinary Research Grant Scheme (Grant No: F07/TRGS/1520/2016).
Authors' contributions
EUHS conceived of and supervised the study. CCS, EUHS and YLC designed the study. CCS performed the data collection and analysis. CCS and EUHS interpreted the results. Both CCS and EUHS contributed to the writing of the manuscript. All authors have read and approved the final manuscript.
Acknowledgements
Not applicable
Table 1: Patient characteristics and clinical information from the 7 relevant studies.
Author |
Year |
Region |
Ethnicity |
Female N (%) |
High/Low Cox-2 N (%) |
Cox-2 detection method |
Threshold/ Cut-off value type |
Patients cancer stages |
Chemotherapy regimen |
Survival outcome |
Li YJ |
2018 |
China |
Asian |
28 (24.3) |
53/62 (46.1/53.9) |
ISH |
NR |
I & II (31), III & IV (84) |
RT, CCRT (platinum) |
OS |
Xu L |
2013 |
China |
Asian |
93 (62.8) |
107/41 (72.3/27.7) |
IHC |
Score |
I & II (95), III & IV (53) |
NR |
OS |
Pan J |
2012 |
China |
Asian |
27 (24.3) |
59/52 (53.2/46.8) |
IHC |
NR |
I & II (30), III & IV (81) |
RT, combined CRT (S-FU) |
OS (10 Yrs) |
Kim TJ |
2011 |
Korea |
Asian |
8 (21) |
31/7 (82/18) |
IHC |
NR |
I & II (6), III & IV (32) |
IC, RT, CCRT |
OS (5 Yrs) |
Huang TL |
2010 |
Taiwan |
Asian |
61 (36) |
NR |
IHC |
H-score |
I & II (71), III & IV (99) |
RT |
OS (5 Yrs) |
Kim TJ |
2010 |
Korea |
Asian |
19 (27.5) |
43/26 (62.3/37.7) |
IHC |
Score |
I & II (17), III & IV (52) |
RT, ICRT, CCRT |
OS |
Loong SL |
2009 |
Singapore |
Asian |
9 (15.5) |
24/34 (41.4/58.6) |
IHC |
Score |
I & II (0), III & IV (58) |
RT, CRT |
OS |
Note: ISH = RNA-in situ hybridization, IHC = Immunohistochemistry, RT = radiotherapy, CCRT = concurrent chemoradiotherapy, IC = induction chemotherapy, ICRT = induction chemoradiotherapy, CRT = chemoradiotherapy, 5-FU = 5-fluorouracil, NR = not related