Elevated Levels of Urinary PGE-M Are Found in Tobacco Users and Indicate a Poor Prognosis for Oral Squamous Cell Carcinoma Patients

Cyclooxygenase (COX) enzymes catalyze the synthesis of prostaglandins from arachidonic acid. In most tissues, COX-1 is constitutively expressed (1). By contrast, COX-2 is induced by both mitogenic and inflammatory stimuli resulting in elevated levels of prostaglandin E₂ (PGE₂) in tumors and inflamed tissues (2–7). COX-2 and PGE₂ play important roles in the development and progression of malignancies. Levels of COX-2 and PGE₂ are increased in both intraepithelial neoplasia and multiple cancers (3–6, 8). The development and growth of experimental tumors are suppressed by treatment with selective COX-2 inhibitors (3, 9–11). Tumor growth is also inhibited in mice engineered to be COX-2 deficient (12–14). In humans, selective COX-2 inhibitors suppressed the formation of sporadic colorectal adenomas (15).

There has been significant interest in the role of COX-2 and PGE₂ in head and neck squamous cell carcinoma (HNSCC), a tobacco-related malignancy (16). Levels of COX-2 are increased in periodontitis, an inflammatory condition that has been associated with an increased risk of oral squamous cell carcinoma (OSCC; refs. 17, 18). Levels of COX-2 are induced by tobacco carcinogens, such as benzo[a]pyrene (19). COX-2 levels are increased in oral leukoplakia, a premalignant condition (9). Moreover, high levels of COX-2 and PGE₂ have been associated with poor prognosis in HNSCC patients (20). Use of NSAIDs, prototypic inhibitors of PGE₂ synthesis, has been associated with a reduced risk of HNSCC in epidemiologic studies (21, 22). Several mechanisms can explain the link between COX-2 and cancer. COX-2–derived PGE₂ can stimulate cell proliferation, migration, and angiogenesis while inhibiting immune surveillance and apoptosis (3, 23–27). PGE₂ has also been suggested to promote metastasis by stimulating cell invasiveness and an epithelial–mesenchymal transition (28, 29).

Because PGE₂ is rapidly catabolized in the lungs, measurements of PGE₂ in plasma do not accurately reflect endogenous production of prostaglandins (30). However, the level of 9, 15-dioxo-11α-hydroxy-13, 14-dihydro-2, 3, 4, 5-tetranor-prostan-1, 20-dioic acid (PGE-M), a stable end metabolite of PGE₂, can be quantified in urine and reflects systemic production of PGE₂ (31, 32). In considering cancer risk factors, levels of urinary PGE-M are increased in both smokers and obese individuals (33, 34). Moreover, several studies have already demonstrated the promise of urinary PGE-M as a cancer risk biomarker (35–37). Increased levels of urinary PGE-M have been found in both non–small cell lung cancer and colorectal cancer patients (38, 39). Finally, in a small retrospective study, we found that high levels of urinary PGE-M were associated with poor prognosis in HNSCC patients (40).

OSCC is the sixth most common cancer worldwide and one of the most common cancers in India. Tobacco quid chewing and smoking are the two most common risk factors for OSCC. In the current study, we had several goals. The first was to determine if levels of urinary PGE-M were increased in healthy individuals who either chewed tobacco quid or smoked cigarettes in the Indian population. The second objective was to determine if levels of urinary PGE-M were increased in OSCC patients. Finally, we investigated whether elevated levels of urinary PGE-M were associated with disease progression in OSCC patients.

Tetranor-PGEM (Catalog No, 14840) and Tetranor-PGEM-d₆ (Item No. 314840) were obtained from Cayman Chemical. Creatinine was obtained from Sigma Aldrich, and creatinine d3 was obtained from Clearsynth Labs. Methanol HPLC grade was obtained from Sigma-Aldrich USA. Acetonitrile (HPLC Grade) and formic acid (GR grade) were purchased from Merck Specialities Pvt. Ltd.; Oasis HLB Cartridge 1CC (30 mg) was obtained from Waters India Ltd. Ultra-pure water of 18 MΩ/cm was obtained from Milli-Q purification system (Millipore).

This study was approved by the Institutional Ethics Committee of Narayana Hrudayalaya. In cohort 1, healthy individuals without a history of malignancy or chronic inflammatory disease were enrolled. After consent, all individuals provided a urine sample and completed a questionnaire regarding tobacco and NSAID use. Individuals with a history of NSAID use within the past week were excluded. The following definitions of tobacco habits were used: no habits, individuals who had never smoked or chewed tobacco/betel quid; current smokers, active smokers with a minimum of a 10 pack year history; quid chewer, individuals who used a minimum of 5 quids/day for 10 or more years but never smoked. We were unable to accrue women who admitted to tobacco smoking.

In cohort 2, we enrolled previously untreated patients with biopsy proven OSCC, for whom treatment with curative intent was planned. After obtaining informed consent, individuals provided a urine sample, and a data collection form was completed prior to surgery. A case–control study design was used. All patients who developed recurrent disease within 2 years of completing treatment were considered cases. Recurrent disease was defined by either pathologic or radiologic evidence of recurrence. Controls were gender, stage, and tobacco habit–matched to cases and were disease free for at least 2 years following completion of treatment. Controls were diagnosed and treated during the same time period as cases. All tumors were staged accordingly to the American Joint Committee on Cancer (AJCC) staging system. Data related to the clinicopathologic features and adjuvant therapy were extracted from the patient records.

A single void urine specimen was collected from each participant and promptly transported to the laboratory after collection. Each sample was aliquoted into 2 mL cryovials and stored at −80°C within 20 minutes of being obtained from the study participant. Samples were stored at −80°C until analysis.

A method for the urinary analysis of tetranor PGE-M as its dehydration product, tetranor PGA-M, in human urine was used as previously described (41). Accordingly, throughout the article, urinary PGE-M will be referred to rather than urinary PGA-M to be consistent with previous publications. In this methods section, we will refer to the compounds by the name (tetranor PGE-M or tetranor PGA-M for clarification).

A different method was used to quantify PGE-M than in our previous studies (33, 34, 40). Due to regulatory requirements, the assays needed to be carried out in India, and this alternate method was chosen (41).

To human urine (1 mL) was added the IS solution [tetranor PGE-M-d₆ (100 ng/mL), 100 μL]. The sample was vortexed for 1 minute. Formic acid (100 μL) was then added to the urine sample to initiate dehydration of tetranor PGE-M to tetranor PGA-M. The sample was mixed and then incubated at 60°C for 24 hours. After incubation, the samples were transferred to a 1 cc/30 mg Oasis HLB SPE column, which had been conditioned with methanol (1 mL), acetonitrile (1 mL) followed by water (1 mL). After application of the samples, the SPE column was dried for 1.0 minute by applying positive pressure at maximum flow rate. The column was washed with 5% acetonitrile in water (1 mL). The samples are eluted with acetonitrile (1 mL). The samples were evaporated to dryness under nitrogen gas and reconstituted in 0.100 mL of 5% acetonitrile in water with 0.1% formic acid. Five microliters were injected in the LC/MS system.

Chromatographic analysis of tetranor PGA-M and tetranor PGA-M-d₆ was achieved on a Agilent 1100 series system (Agilent Technologies) under reverse-phase conditions. Separation of tetranor PGA-M and tetranor PGA-M-d₆ was performed on Xterra C18 (Waters Corporation; 5 μm, 4.6 × 250 mm) that was maintained at 30°C in a column oven. A gradient mobile phase consisting of 0.1% formic acid in water (A) and 0.1% formic acid in acetonitrile (B; flow rate 0.750 mL/min with 70% flow split) was used to separate the analytes (minutes, % mobile phase A): (0, 95), (0.2, 95), (0.21, 85), (1.00, 85), (1.10, 70), (6.00, 70), (6.10, 5), (8.00, 5), (8.10, 95), and (14.00, 95). Analysis of tetranor PGA-M and tetranor PGA-M-d₆ was carried out on a triple quadrupole mass spectrometer (API 4000 MDS Sciex; Applied Biosystems USA), equipped with Turbo Ion Spray ionization and operated in a Negative ionization mode. The MS/MS detection monitored the transition m/z 309.00→143.00 for tetranor PGA-M and m/z 315.00→149.10 for tetranor PGA-M-d₆, and the ratios of the peak areas for each of these analytes were used for quantitation. Peak integration and calibration were carried out using Analyst 1.4.2 software.

Creatinine concentrations were quantified using previously established methods (41).

5 mL of mobile phase containing the internal standard (creatinine–d₃, 10 μg/mL) was added to an aliquot of 5 μL human urine and mixed for 1 minute. Mixture was centrifuged at 4,500 rpm at 10°C for 5 minutes. Supernatant (5 μL) was injected in the system.

Chromatographic analysis of creatinine was achieved on an Agilent 1100 series system (Agilent Technologies) under reverse-phase conditions. Separation of creatinine and creatinine–d₃ was performed on a Zorbax Eclipse XDB-C18 (Agilent Technologies; 3.5 μm, 100 × 4.6 mm) and maintained at 30°C in a column oven. Separation was achieved using isocratic analysis with a mobile phase of water: acetonitrile (40:60, v/v) with 0.1% formic acid and a flow rate of 1.00 mL/min with 70% flow split. Analysis of creatinine and creatinine–d₃ was carried out on a triple quadrupole mass spectrometer (API 4000 MDS Sciex; Applied Biosystems USA), equipped with Turbo Ion Spray ionization and operated in a positive ionization mode. The MS/MS detection monitored the transition m/z 114.00→86.00 for creatinine and m/z 117.00→ 47.20 for creatinine–d₃, and the ratios of the peak areas for each of these analytes were used for quantitation. Peak integration and calibration were carried out using Analyst 1.4.2 software.

Characteristics of healthy subjects and clinicopathologic variables in OSCC patients were summarized in terms of median (range) for continuous variables and count (proportion) for categorical variables. Differences in the values of a continuous variable across more than two groups and between two groups were examined using the nonparametric Kruskal–Wallis test and Wilcoxon rank-sum test, respectively. Differences in the distribution of a categorical variable across/between study groups were examined using the Fisher exact test. Time-to-event data, including progression-free survival and overall survival, for subjects with high (above median) and low (below median) urinary PGE-M levels are summarized using Kaplan–Meier curves. The log-rank test was used to examine the association between each of the independent variables and a time-to-event outcome univariably. The Cox proportional hazards model was used for multivariable analysis of the association between urinary PGE-M and the outcome adjusting for other covariates, including those demonstrated plausible association with P value < 0.25 in the univariate analyses and the known prognostic factors. The HRs with 95% confidence interval (CI) and P values were reported. All tests were two sided, and P values of <0.05 were considered statistically significant.

Initially, we investigated whether levels of urinary PGE-M were altered in healthy tobacco quid chewers or smokers. A total of 94 subjects were included: 30 never tobacco quid chewers-never smokers, 33 tobacco quid chewers, and 31 smokers (Table 1). Participants in these three groups were matched for age and body mass index (BMI). Overall, a statistically significant elevation in median urinary PGE-M levels was found in healthy tobacco quid users (21.3 ng/mg Cr; P = 0.03) and smokers (32.1 ng/mg Cr; P < 0.001) compared with never tobacco quid chewers-never smokers (18.8 ng/mg Cr; Fig. 1A). Next, we determined whether this effect on urinary PGE-M varied by gender. Male smokers (P < 0.001) and tobacco quid chewers (P = 0.03) had higher levels of PGE-M than male never tobacco quid chewers-never smokers (Fig. 1B). For females who chewed tobacco quid compared with never tobacco quid chewers-never smokers, there was a trend toward higher urinary PGE-M levels, but this difference did not reach statistical significance (Fig. 1C). Interestingly, the amount of tobacco quid use was slightly greater in females (median, 120 quid years) than in males (median, 90 quid years), although this difference was not statistically different.

Next, we investigated whether levels of urinary PGE-M were elevated in OSCC patients compared with healthy controls who chewed tobacco quid or smoked. Seventy-eight previously untreated patients with biopsy proven OSCC undergoing curative intent treatment were included. The OSCC patient characteristics are shown in Table 2. Several differences were observed between the OSCC patients (Table 2) and healthy tobacco users (Table 1). The OSCC patients were older and had a lower BMI than healthy controls (P < 0.001). All OSCC patients were either smokers (35.9%) or tobacco quid chewers (64.1%). The median tobacco pack year exposure for healthy tobacco smokers was 12 compared with 6.6 for OSCC patients who smoked (P < 0.001). Quid years were calculated based on the number of tobacco quids used per day. For tobacco quid users, the median quid years was significantly higher among OSCC patients (200) than healthy quid users (96; P = 0.003). Levels of urinary PGE-M were compared in 78 OSCC patients versus 64 healthy tobacco users. An approximately 1-fold increase in the median level of urinary PGE-M was found in OSCC patients (48.7 ng/mg Cr) versus healthy tobacco users (24.5 ng/mg Cr; P < 0.001). In a multivariable analysis that controlled for age, gender, and BMI, the difference in levels of urinary PGE-M between OSCC patients and healthy tobacco user controls remained significant (P = 0.004). Notably, the increase in urinary PGE-M was found in both male (Fig. 2A) and female (Fig. 2B) OSCC patients compared with healthy (normal) tobacco users. Moreover, the increased levels of urinary PGE-M in OSCC patients were detected in both smokers (Fig. 2C) and tobacco quid chewers (Fig. 2D).

We next investigated the prognostic significance of urinary PGE-M levels in OSCC patients. As shown in Table 2, patients were grouped as those who either progressed (n = 37) or remained progression free (n = 41). The patients who had local recurrence, regional, or distant metastasis following attempted curative treatment were grouped as having progressive disease. None of the patients in this study had clinical evidence of second primary tumors. Age at the time of OSCC diagnosis and tobacco habits were not significantly different. The majority of patients in both groups had stage III or IV disease at presentation. Patients who developed progressive disease had relatively higher nodal stage (N2, 32.4%) compared with the progression-free group (N2, 14.6%; P = 0.19). Patients who progressed had a greater proportion of poorly differentiated tumors than patients who did not progress, although the differences did not reach statistical significance (P = 0.24). Other pathologic features (margin status, perineural invasion, lymphovascular invasion, and extracapsular spread) were statistically similar in the two groups.

Baseline levels of urinary PGE-M were compared in the two groups (progression vs. progression free) of OSCC patients. A nearly 1-fold (64.7 vs. 33.8 ng/mg Cr, P < 0.001) higher median baseline level of urinary PGE-M was found in the group of OSCC patients who progressed compared with the group that remained progression free (Table 2; Fig. 3). Notably, there was considerable variability in the levels of urinary PGE-M within each of the two groups. In addition, higher urinary PGE-M levels were associated with a significantly increased risk of recurrence (HR, 1.01 per unit increase; 95% CI, 1.01–1.02; P < 0.001). Dichotomizing urinary PGE-M at the median (48.7 ng/mg Cr), patients with high urinary PGE-M (above median) had significantly increased risk of progression compared with patients with low urinary PGE-M (HR, 2.20; 95% CI, 1.13–4.29; P = 0.02; Fig. 4A). Similarly, we examined the association between urinary PGE-M levels and overall patient survival. Higher urinary PGE-M levels were significantly associated with worse overall survival (HR, 1.01 per unit increase; 95% CI, 1.00–1.02; P = 0.003). Dichotomizing urinary PGE-M at the median, patients with high urinary PGE-M (above median) had a trend toward poorer overall survival (HR, 1.93; 95% CI, 0.95–3.90; P = 0.06; Fig. 4B).

We chose variables with P values less than 0.25 in the univariate analysis to be the independent variables in a multivariable model. In a multivariable model that controlled for age, BMI, gender, tobacco use, tumor differentiation, presence of extracapsular spread, and adjuvant chemotherapy, elevated baseline urinary PGE-M was associated with increased risk of progression (HR, 1.01 per unit increase; 95% CI, 1.01–1.02; P < 0.001). For overall survival, we also examined the association between each covariate using the Cox proportional hazards model. Multivariable analysis suggested that high urinary PGE-M remained significantly associated with increased risk of death (HR, 1.01 per unit increase; 95% CI, 1.00–1.02; P = 0.03) after adjusting for age, BMI, gender, tobacco use, AJCC stage, tumor differentiation, extracapsular spread, and adjuvant chemoradiation.

In the current study, we showed a link between tobacco use and increased levels of urinary PGE-M. Elevated urinary PGE-M levels were found in healthy male tobacco quid chewers. By contrast, a nonstatistically significant increase in urinary PGE-M was observed in healthy female tobacco quid chewers. One likely explanation for the observed increase in urinary PGE-M in men is tobacco quid–related periodontitis or other inflammatory conditions in the oral cavity (42). COX-2 is overexpressed in periodontitis (17). Notably, some studies have suggested a link between periodontitis and increased risk of OSCC (18). Possibly, the observed elevation in urinary PGE-M reflects an increased risk of OSCC in tobacco quid chewers. Why the effect of tobacco quid on urinary PGE-M levels was less dramatic in women is uncertain. Based on history alone, we cannot exclude the possibility that tobacco quid exposure in women was less than in men. In this context, we also note that estrogen possesses anti-inflammatory properties (43). It is possible, therefore, that estrogen antagonizes the induction of COX-2 in tobacco quid chewers leading to less elevation in urinary PGE-M. The current results also confirm our original finding that levels of urinary PGE-M are increased in smokers versus never smokers (33). Notably, we previously demonstrated that the increased levels of urinary PGE-M in smokers reflected increased COX-2 activity (33). In all likelihood, the lung is the source of the increased COX-2 expression and PGE₂ synthesis in smokers because of the known link between smoking and lung inflammation. In support of this possibility, increased concentrations of PGE₂ have been observed in the sputum of both smokers and patients with COPD compared with nonsmoking controls (44). The fact that urinary PGE-M levels were significantly higher in smokers compared with tobacco quid chewers is consistent with inflammatory effects in the lung which has a very large surface area versus local effects in the oral cavity.

We also found increased urinary PGE-M in OSCC patients in comparison with healthy tobacco users. Previously, we found a near statistically significant increase in urinary PGE-M in HNSCC patients (45). The fact that the effects were more dramatic in the current study is likely to reflect differences in study design. The current study was larger than the original one and limited to patients with OSCC. Moreover, in the original HNSCC study, the patient population was more heterogeneous and included multiple tumor sites including the larynx and paranasal sinuses (45). The observed elevation in urinary PGE-M fits with the known elevation of PGE₂ in HNSCC (5, 16). The fact that levels of COX-2 and microsomal prostaglandin E synthase-1, the enzyme that converts COX-2–derived PGH₂ to PGE₂, are commonly overexpressed in HNSCC can explain both the elevated levels of intratumoral PGE₂ and urinary PGE-M (8, 46). Notably, the elevation in urinary PGE-M levels in OSCC patients is consistent with prior evidence of increased urinary PGE-M levels in patients with other malignancies (38, 39).

Previously, we carried out a small retrospective study in HNSCC patients which suggested that elevated urinary PGE-M levels were associated with more rapid disease progression and reduced overall survival (40). In fact, the current larger prospective study was carried out, in part, to validate and extend upon these earlier findings. Here, we demonstrated that elevated levels of urinary PGE-M were associated with reduced disease-specific and overall survival. These findings are also consistent with prior studies which reported that both elevated COX-2 and PGE₂ at the invasive front of the tumor predict for worse disease-free and overall survival in HNSCC patients (20, 47). As mentioned above, PGE₂ has multiple effects that can potentially explain the link between elevated urinary PGE-M and poor prognosis. PGE₂ exerts its effects by binding to anyone of four plasma membrane receptors known as EP receptors in stromal or tumor cells leading to increased cell proliferation, reduced apoptosis, and enhanced angiogenesis (3). Cross-talk between EP receptors and the EGFR can also stimulate cell growth (3). PGE₂ has multiple other effects that may stimulate tumor progression including induction of matrix metalloproteinase-9 and VEGF and stimulation of epithelial–mesenchymal transition (3, 29, 48). In addition, there is considerable evidence that COX-derived PGE₂ is a driver of tumor growth through evasion of immunity (49). We were unable to assess the relationship between urinary PGE-M levels at the time of diagnosis and the risk of developing second primary tumors. Future studies are warranted to address this question.

A major question that arises based on the current results is whether OSCC patients with high urinary PGE-M levels would benefit from adjuvant treatment with traditional NSAIDs, a selective COX-2 inhibitor, or aspirin. Several studies have demonstrated that traditional NSAIDs, selective COX-2 inhibitors, and aspirin suppress urinary PGE-M levels (33, 38, 41). In preclinical models, inhibitors of PGE₂ production have proven useful for preventing or treating a variety of tumors, including HNSCC (3, 11). In a country such as India where OSCC is extremely common, inexpensive medicines such as these, if active, would be very beneficial. Certainly, our results underscore the need for future studies to determine whether COX inhibitors can be used in the treatment of OSCC or perhaps to reduce the risk of second primaries.

C. Kandasamy is Head—Bioanalytical at Syngene International Ltd. No potential conflicts of interest were disclosed by the other authors.

Conception and design: V.D. Kekatpure, X.K. Zhou, A.J. Dannenberg

Development of methodology: V.D. Kekatpure

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): V.D. Kekatpure, Naveen BS, C. Kandasamy, S.P. Sunny, A. Suresh, M.A. Kuriakose

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): V.D. Kekatpure, H. Wang, X.K. Zhou, G.L. Milne, A.J. Dannenberg

Writing, review, and/or revision of the manuscript: V.D. Kekatpure, X.K. Zhou, A. Suresh, G.L. Milne, M.A. Kuriakose, A.J. Dannenberg

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): V.D. Kekatpure, X.K. Zhou, S.P. Sunny

This study was supported by a research grant from the Biocon Foundation (to V.D. Kekatpure).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.