0
Original Research: Lung Cancer |

Serum Free Fatty Acid Biomarkers of Lung CancerFatty Acid Lung Cancer Biomarkers FREE TO VIEW

Jinbo Liu, MD; Peter J. Mazzone, MD, MPH; Juan P. Cata, MD; Andrea Kurz, MD; Maria Bauer, MD; Edward J. Mascha, PhD; Daniel I. Sessler, MD
Author and Funding Information

From the Department of Outcomes Research (Drs Liu, Cata, Kurz, Bauer, Mascha, and Sessler), the Department of Pulmonary, Allergy, and Critical Care Medicine (Dr Mazzone), the Department of Quantitative Health Sciences (Dr Mascha), and the Cleveland Clinic Lerner College of Medicine (Dr Liu), Cleveland Clinic, Cleveland, OH; and the Department of Anesthesiology and Perioperative Medicine (Dr Cata), University of Texas MD Anderson Cancer Center, Houston, TX.

CORRESPONDENCE TO: Daniel I. Sessler, MD, Department of Outcomes Research, Cleveland Clinic, 9500 Euclid Ave—P77, Cleveland, OH 44195; e-mail: DS@OR.org


FUNDING/SUPPORT: This study was funded in part by the Joseph Drown Foundation (Los Angeles, CA).

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details.


Chest. 2014;146(3):670-679. doi:10.1378/chest.13-2568
Text Size: A A A
Published online

BACKGROUND:  Lung cancer is the leading cause of cancer-related mortality. Surgical removal of the tumor at an early stage can be curative. However, lung cancer diagnosis at an early stage remains challenging. There is evidence that free fatty acids play a role in cancer development.

METHODS:  Serum samples from 55 patients with lung cancer were propensity matched with samples from 165 similar pulmonary patients without known cancer. Patients were propensity matched on age, sex, smoking history, family history of lung cancer, and chronic diseases that might affect free fatty acid levels.

RESULTS:  Free fatty acids arachidonic acid (AA) and linoleic acid (LA) and their metabolites hydroxyeicosatetraenoic acids (HETEs)(5-HETE, 11-HETE, 12-HETE, and 15-HETE) were an estimated 1.8- to 3.3-fold greater in 37 patients with adenocarcinoma vs 111 patients without cancer (all P < .001). Areas under the receiver operating characteristic curve were significantly > 0.50, discriminating patients with lung cancer and control subjects for six of eight biomarkers and two of seven phospholipids tested, and ranged between 0.69 and 0.82 (all P < .001) for patients with lung cancer vs control subjects. AA, LA, and 15-HETE had observed sensitivity and specificity > 0.70 at the best cutpoint. Concentrations of free fatty acids and their metabolites were similar in 18 patients with squamous cell carcinoma and 54 control subjects without cancer.

CONCLUSIONS:  Serum fatty acids and their metabolites demonstrate good sensitivity and specificity for the identification of adenocarcinoma of the lung.

Figures in this Article

Lung cancer is among the most common types of cancer. Surgery is curative only at relatively early stages. By the time lung cancer becomes symptomatic, the best chance for surgical cure is usually well past. Consequently, lethality is high, and lung cancer remains the leading cause of cancer deaths in the United States and worldwide.

Chest radiographs are not sensitive for detection of early-stage lung cancer, and yearly chest radiographs do not reduce lung cancer mortality.1,2 The only currently accepted method of screening for lung cancer is low-dose CT imaging.3,4 However, CT imaging is expensive and exposes patients to small doses of radiation. Furthermore, approximately 95% of lung nodules identified by CT scan are benign, and it can be challenging for radiologists to identify those that are not.4 Low-dose CT scan screening for lung cancer could, thus, be optimized with tools that improve risk stratification and nodule management.

Fatty acids and phospholipids are necessary for cancer cell proliferation.5 Platelet activating factor (PAF), a potent phospholipid, promotes lung cancer growth and metastasis.6 PAF is hydrolyzed by phospholipase A2 to produce lysophospholipid PAFs (lysoPAFs). Patients with lung cancer have increased expression and activity of phospholipase A2.7,8 This enzyme catalyzes oxidized phospholipids to form lysoPAFs and free fatty acids (FFAs), which, in turn, play a major role in tumor development, angiogenesis, lymphangiogenesis, progression, and metastasis.913 LysoPAFs show altered expression in cancer cells. They bind to and activate specific cell-surface G protein-coupled receptors that initiate cell growth, proliferation, and survival pathways (Fig 1).14,15 LysoPAFs (eg, lysophosphatidylcholines) have been proposed as a potential biomarker for ovarian cancer.16,17 Plasma phospholipids, including phosphatidylcholine and lysophosphatidylcholines, may be potential biomarkers in prostate cancer18 and lung cancer.19

Figure Jump LinkFigure 1  The role of LPs in tumor development. LP binds its cognate GPCR and stimulates cell proliferation and migration, which may initiate cancer cell development and metastasis. GPCR = G-protein-coupled receptor; LP = lysophospholipid.Grahic Jump Location

The theory that elevated FFAs promote cancer is supported by the observation that high dietary linoleic acid (LA) intake promotes breast cancer metastasis in animals.20,21 The key enzymes for fatty acid synthesis are upregulated in cancer cells, play a critical role in cancer development, and have been chosen as anticancer therapeutic targets.2226 The hydroxyeicosatetraenoic acids (HETEs) and hydroxyoctadecadienoic acids (HODEs) are stable oxidative products of arachidonic acid (AA) and LA, respectively. There is considerable basic science evidence that FFAs and the oxidized FFAs HETE and HODE promote tumor development, progression, and metastases in lung cancer using cell culture,27,28 tissues,2931 and animal models.27,3234 For example, 5-HETE augments lung cancer cell survival,30,35 12-HETE enhances lung cancer cell adhesion,3639 and 15-HETE promotes human lung cancer metastasis40 and induces angiogenesis.41,42 HODE is involved in many types of cancer, although its specific role remains unclear.4346 Studies show that 9-HODE and 13-HODE are ligands of peroxisome proliferator-activated receptor-γ, which promote lung cancer progression and metastasis.47,48 We, therefore, tested the hypotheses that serum concentrations of phospholipids, FFAs, and their metabolites are greater in patients with known lung cancer than in matched patients apparently without cancer.

With Cleveland Clinic Institutional Review Board approval (IRB # 13-306) and informed consent, we retrieved serum samples from 55 patients with lung cancer (cases) from the Cleveland Clinic Lung Biobank. Serum samples were collected from patients with lung cancer at the time of their diagnosis, prior to the initiation of treatment. Pathologic analysis indicated that 37 had lung adenocarcinoma and 18 had squamous cell carcinoma.

Control subjects were selected from a biobank whose inclusion criteria included age 40 to 75 years and at least one of the following criteria: (1) current or ex-smoker with at least a 10 pack-year history, (2) a first-degree relative with a history of lung cancer, or (3) a clinical diagnosis of COPD. Specific matched control subjects (three for each patient with cancer) were chosen separately for patients with adenocarcinoma and squamous cell carcinoma from among all available control patients (see statistical analysis).

Sample Processing

Samples were centrifuged at 3,000 × g within 2 h of collection, and the resulting serum was frozen at −70°C until assayed. Biochemical analysis was conducted by an investigator who was strictly blinded to cancer status and patient characteristics.

Two hundred microliters of serum from each patient was used for phospholipid and FFA extraction. To each of the serum samples we added an equal amount of PAF (1-O-hexadecyl-2-acetyl-sn-glycero-3-phosphocholine)-d4 and 15-HETE-d8 (Cayman Chemical Company) as internal standards for phospholipids or FFAs, respectively. The phospholipids and FFAs were extracted first using chloroform/methanol and then further isolated by column chromatography. Fractions of phospholipids and FFA were dried by liquid nitrogen and dissolved in 200 μL of 85% menthol or 50% methanol in high-performance liquid chromatography water, respectively.

Mass spectrometric analyses for phospholipids were performed online using electrospray ionization tandem mass spectrometry in the positive ion mode with multiple reaction monitoring using the molecular cation [MH]+ and the m/z 184 daughter phosphocholine ion. For FFA analysis, negative-ion mode with multiple reaction monitoring of parent and individual daughter ions of oxidized and unoxidized fatty acids used the m/z transitions: 5-HETE (319→115); 8-HETE (319→155); 9-HETE (319→151); 11-HETE (319→167); 12-HETE (319→179); 15-HETE (319→175); 20-HETE (319→245); 9-HODE (295→171); 13-HODE (295→195); arachidonate (303→259); and linoleate (279→261).

Statistical Analysis

Specific control subjects (three for each patient with cancer) were chosen separately for patients with adenocarcinoma and squamous cell carcinoma from among all available control patients using propensity score matching. For each cancer type, a logistic regression model predicting cancer status (yes or no) from all available baseline variables was constructed, and an estimated propensity score estimated for each patient. Control patients were matched in a 3:1 ratio to the patients with cancer on the propensity score to within 0.2 SDs of the distribution of the logit of the propensity score using a greedy matching algorithm.

Factors included in each propensity score model, regardless of their statistical significance, were sex, age, smoking history (former/never, current), a diagnosis of COPD, a family history of lung cancer, and history of diabetes. For each cancer type, balance on baseline variables between matched patients with cancer and control patients was assessed using the standardized difference (ie, the difference in means or proportions divided by the pooled SD). For patients with cancer, we also assessed the correlation between stage of cancer and each potential biomarker.

We assessed the diagnostic accuracy of the FFAs and their metabolites for detecting lung cancer using receiver operating characteristic analysis and estimating the area under the curve (AUC) with CI. We searched for best cutpoints for maximizing sensitivity and specificity with each compound and set a priori that a clinically reliable cutpoint would require observed sensitivity and specificity of at least 0.70 (arguably, at least 0.90 would be needed before a marker would be used in practice; in this exploratory analysis we purposely set a low bar). For each cancer type we also assessed the univariable and multivariable association between the potential biomarkers and presence of cancer using logistic regression, with results expressed as OR with CIs and overall AUC.

To control the type 1 error at 5% across the assessment of potential biomarkers within each of the two cancer types, we deemed P values significant if P < .05/15 = .0033, adjusting for six FFAs, seven phospholipids, and the variables LA and AA. With a total of 37 patients with adenocarcinoma and 111 control subjects, we had 90% power at the overall .05 significance level to detect differences as small as 0.82 SDs between patients with cancer and control patients on a particular metabolite. With N = 18 patients with squamous cell carcinoma and 54 control subjects we had 90% power to detect differences as small as 1.2 SDs. SAS statistical software (SAS Institute Inc) or the R programming language was used for all analyses.

Propensity matching of the registry patients resulted in well-balanced cancer and control groups for both adenocarcinoma and squamous cell carcinoma, evidenced by nonsignificant P values and relatively small standardized differences on baseline variables (Tables 1, 2). Three-quarters of the patients with adenocarcinoma had stage 3 or 4 disease. For patients with adenocarcinoma, the Spearman correlation with stage of cancer was near zero (< 0.10) for eight of 16 potential biomarkers and between 0.16 and 0.42 for the other eight (P < .05 for 11_HETE, 12-HETE, and15-HETE). For patients with squamous cell carcinoma, stage was not significantly correlated with any of the potential biomarkers, with correlations ranging from 0 to 0.30. Unmatched control subjects were somewhat more likely to be men (51%), younger (mean age 61 years), and have lower pack-years of smoking (median 34) and less likely to have COPD (16%) or diabetes mellitus (8%) than those matched to the patients with cancer.

Table Graphic Jump Location
TABLE 1  ] Matched Patients With Adenocarcinoma and Control Patients: Baseline Characteristics

Results presented as No. (%) unless otherwise noted. DM = diabetes mellitus; F = Fisher exact test; STD = standardized difference; W = Wilcoxon rank sum test.

a 

Pearson χ2 test, unless otherwise noted.

b 

Difference in means or proportions/pooled SD.

Table Graphic Jump Location
TABLE 2  ] Matched Patients With Squamous Cell Carcinoma and Control Patients: Baseline Characteristics

Results presented as No. (%) unless otherwise noted. See Table 1 legend for expansion of abbreviations.

a 

Pearson χ2 test, unless otherwise noted.

b 

Difference in means or proportions/pooled SD.

Ratio of Means

In patients with adenocarcinoma, serum oxidized FFAs 5-HETE, 11-HETE, 12-HETE, and 15-HETE ranged an estimated 1.8- to 3.3-fold greater in patients with lung cancer than in patients without cancer (Fig 2, Table 3) (P < .001), whereas LA and AA were, respectively, an estimated 1.7 (99.7% CI, 1.2-2.3)-fold and 1.8 (1.4-2.3)-fold greater in patients with cancer than control subjects. No difference was found for either 9-HODE (P = .79) or 13-HODE (P = .56). 8-HETE, 9-HETE, and 20-HETE were undetectable in most samples.

Figure Jump LinkFigure 2  Box plots for predictors of adenocarcinoma in patients with and without cancer. Displayed predictors needed to have area under the receiver operating characteristics curve significantly > 0.50 and both sensitivity and specificity > 0.70 at a “best” cutpoint, which maximizes both parameters. Top and bottom of box are first and third quartiles, respectively. Solid line in side box is median, diamond represents the mean. Whiskers extend to the minimum of 1.5 interquartile ranges or the range of the data. AA = arachidonic acid; CA = cancer; HETE = hydroxyeicosatetraenoic acid; LA = linoleic acid.Grahic Jump Location
Table Graphic Jump Location
TABLE 3  ] Adenocarcinoma: Comparing Patients With Cancer and Control Patients on Fatty Acids and Phospholipids

AA = arachidonic acid; AzPAF = azelaoyl platelet activating factor; AzPC = azelaoyl phosphatidylcholine; HETE = hydroxyeicosatetraenoic acid; HODE = hydroxyoctadecadienoic acid; HODE-PC = hydroxyoctadecadienoylphosphatidylcholine; HpODE-PC =hydroperoxyoctadecadienoyl phosphatidylcholine; LA = linoleic acid; LPC = lyso-phosphatidylcholine; lysoPAF = lysophospholipid platelet activating factor; Q1 = first quartile = 25th percentile; Q3 = third quartile = 75th percentile. See Table 1 legend for expansion of other abbreviation.

a 

Estimated ratio of geometric means, calculated as exponentiated difference in means of the log-transformed data; 99.7% CI maintains type 1 error at 5% across 15 variables.

b 

Two-tailed Wilcoxon rank sum test.

c 

Significant if P < .05/15 = .0033.

Patients with adenocarcinoma had lower concentrations of oxidized phospholipid hydroxyoctadecadienoylphosphatidylcholine (HODE-PC) and 1-O-Hexadecyl-sn-glycero-3-phosphocholine compared with control subjects without cancer, but no difference was found for the other five phospholipids assessed (P > .0033), including C16-lysoPAF, C18-lysoPAF, and oxidized phospholipid hexadecylazelaoylphosphatidylcholine (Table 3). For patients with squamous cell carcinoma, none of the FFAs or their metabolites or phospholipids had higher means in patients with cancer vs control patients after Bonferroni correction (Table 4).

Table Graphic Jump Location
TABLE 4  ] Squamous Cell: Comparing Patients With Cancer and Control Patients on Fatty Acid and Phospholipids

See Table 1 and 3 legends for expansion of abbreviations.

a 

Estimated ratio of geometric means, calculated as exponentiated difference in means of the log-transformed data; 99.7% CI maintains type 1 error at 5% across 15 variables.

b 

Two-tailed Wilcoxon rank sum test.

c 

Difference in means or proportions/pooled SD.

Diagnostic Accuracy

Tables 5 and 6 report the estimated area under the receiver operating characteristic curve and diagnostic accuracy results for all biomarkers in which the AUC was significantly > 0.50 after Bonferroni correction (P < .0033). For patients with adenocarcinoma, estimated AUC ranged from 0.71 to 0.82 for metabolites 5-HETE, 11-HETE, 12-HETE, and 15-HETE as well as AA and LA (Table 5). Estimated AUC was 0.47 (P = .59) for 13-HODE and 0.49 (P = .80) for 9-HODE (not shown). In a multivariable logistic regression, AUC increased to 0.87 for patients with adenocarcinoma vs control patients when including any of the biomarkers in Table 5 significant at .05 in the presence of the others (the final backward selection model included 12-HETE, 15-HETE, 9-HODE, and LA).

Table Graphic Jump Location
TABLE 5  ] Adenocarcinoma: Diagnostic Accuracy of Each FFA Metabolite Predicting Lung Cancer

N = 37 patients with cancer, 111 control subjects. Diagnostic accuracy parameters (sensitivity, specificity, and so forth) correspond to the given threshold value. CI estimated using bootstrap resampling (percentile method) with 5,000 resamples. Accuracy: ratio of sum of true positives, true negatives, false positives, false negatives to total sample size. Data values represent peak raw biomarker values divided by the corresponding internal standard for each sample. N = 3 biomarkers met the criteria of having sensitivity and specificity at the best cutpoint > 0.70 (AA, LA, HETE15). CI: 99.7% CIs adjusted for assessing multiple biomarkers using Bonferroni correction; overall α = 0.05. FFA metabolite values are metabolite peak/internal standard peak. AUC = area under the curve; FFA = free fatty acid; NPV = negative predictive value; PPV = positive predictive value. See Table 3 legend for expansion of other abbreviations.

a 

All P values < .001 except for AzPAF (P = .027), HpODE-PC (P = .040).

b 

Jointly maximizes sensitivity and specificity.

c 

Estimate using Bayes’ theorem assuming a true prevalence of 0.10.

Table Graphic Jump Location
TABLE 6  ] Squamous Cell: Diagnostic Accuracy of Each FFA Metabolite Predicting Lung Cancer

N = 18 patients with cancer, 54 control subjects. FFA metabolite values are metabolite peak/internal standard peak. Diagnostic accuracy parameters (sensitivity, specificity, etc.) correspond to the given threshold value. CI estimated using bootstrap resampling (percentile method) with 5,000 resamples. Accuracy: ratio of sum of true positives, true negatives, false positives, false negatives to total sample size. Data values represent peak raw biomarker values divided by the corresponding internal standard for each sample. No biomarkers met the criteria of having sensitivity and specificity at the best cutpoint > 0.70. CI: 99.67% CIs adjusted for assessing multiple biomarkers using Bonferroni correction; overall α = 0.05 CI: 99.7% CIs adjusted for assessing multiple biomarkers using Bonferroni correction; overall α = 0.05. See Table 3 and 5 legends for expansion of abbreviations.

a 

Jointly maximizes sensitivity and specificity.

b 

Estimate using Bayes’ theorem assuming a true prevalence of 0.10. AUC P value: testing H0: AUC = 0.5 (1-tailed).

For the adenocarcinoma biomarkers with statistically significant AUC (n = 8 of 15 tested), estimated sensitivity ranged from 0.65 to 0.73 and estimated specificity ranged from 0.65 to 0.74 at the thresholds, which maximized sensitivity and specificity (Table 5). Estimates of accuracy at the chosen thresholds ranged from 0.65 to 0.76. However, only 15-HETE, AA, and LA met our a priori “usefulness” criterion of having both observed sensitivity and observed specificity > 0.70 at the chosen threshold. Figure 3 plots the receiver operating characteristic curve of sensitivity vs 1 minus specificity at each observed value for these three candidate biomarkers.

Figure Jump LinkFigure 3  Receiver operating characteristic curves with area under the curve (AUC) and SE in parentheses for patients with adenocarcinoma. Displayed are the three variables that had AUC significantly > 0.50 and both sensitivity and specificity estimated as > 0.70 (see Table 5). See Figure 2 legend for expansion of abbreviations.Grahic Jump Location

In patients with squamous cell carcinoma, the AUC was significantly > 0.50 in the four biomarkers/phospholipids shown in Table 6: AA, LA, 11-HETE, and lysoPAF. Of those, estimated sensitivity and specificity was > 0.70 only for lysoPAF.

Serum concentrations of FFAs and their oxidized metabolites (HETEs and HODEs) were substantially greater in patients with lung adenocarcinoma than in similar patients without cancer. Among the potential biomarkers we evaluated, AA, LA, and 15-HETE showed the best combination of sensitivity and specificity (> 0.70) and AUC (range, 0.76-0.82).

Predictive value for adenocarcinoma was improved by considering combinations of FFAs and metabolites in a multivariable model, with a combined AUC of 0.87. These results suggest that these serum molecules, or a combination of them, may be good biomarkers for lung adenocarcinoma diagnosis.

In contrast to adenocarcinoma, in patients with squamous cell cancer, only C18-lysoPAF passed our criterion of significant AUC and observed sensitivity/specificity > 0.70. The AUC for that molecule was only 0.72 (99.6% CI, 0.52-0.91). Because just 18 patients had squamous cell cancer, we only had 90% power to detect differences of 1.2 SDs. FFAs and their oxidized metabolites thus appear to better detect adenocarcinoma than squamous cell cancer.

Many types of molecular biomarkers are being developed to assist with lung cancer diagnosis and management. One advantage of the development of a biomarker based on FFAs and their oxidized metabolites is that they are stable. FFA concentrations in whole blood remain unchanged up to 28 h at room temperature,49 up to 12 months at −20°C,5052 and up to 10 years at −80°C.53 Stability of these molecules makes them clinically practical biomarkers, because special specimen handling is unnecessary.

These patients with lung cancer and control patients were matched on age, sex, smoking status, family history of lung cancer, and history of diseases such as diabetes and COPD. Furthermore, both groups were selected from a similar population seen in the Cleveland Clinic Pulmonary Clinic. Although it remains probable that the cancer and control populations differed on some unknown but important factors, the very substantial differences in the FFA-based biomarker concentrations is most likely to result from the metabolic changes associated with the presence of lung cancer. Lung cancer was the only tumor we evaluated; whether FFA, HETE, and HODE are comparably elevated in patients with other types of cancer remains unknown.

The current report reflects the results of a discovery-level project. Much additional work is required to determine the potential clinical usefulness of an FFA-based lung cancer serum biomarker. Our testing set included all stages of lung cancer. To develop the biomarker for early identification will require validation of the results within an independent cohort with a focus on early-stage lung cancers. Characterizing true and false results will help to determine its potential clinical application. The testing platform used must be translated into a laboratory-ready platform and validated to deliver accurate results consistently. Potential applications of a validated FFA-based lung cancer serum biomarker that might be explored include identification of high-risk patients for enrollment in lung cancer screening programs, characterization of lung nodules, monitoring the response to therapy, and characterizing the nature of an individual’s cancer.

In summary, serum FFAs and their oxidized metabolites, HETE and HODE, were elevated in patients with lung adenocarcinoma. Among these, AA and LA and 15-HETE were best able to distinguish cancer from noncancer status measured by the AUC and also had the best sensitivity and specificity. Future work could identify a role for these compounds in the identification and characterization of lung cancer.

Author contributions: J. L. and D. I. S. are guarantors of the article. D. I. S. acts as archival author. J. L., P. J. M., and D. I. S. contributed to study concept; J. L., P. J. M., A. K., E. J. M., and D. I. S. contributed to study design; E. J. M. contributed to study analysis; J. L. contributed to bioanalysis of patient serum samples; P. J. M., J. P. C., and M. B. contributed to data acquisition; J. L., P. J. M., E. J. M., and D. I. S. contributed to manuscript preparation; and J. L., P. J. M., J. P. C., A. K., M. B., E. J. M., and D. I. S. contributed to approval of the final version of the manuscript.

Financial/nonfinancial disclosures: The authors have reported to CHEST the following conflicts of interest: Drs Liu and Sessler have intellectual property related to this report. Dr Mazzone has intellectual property related to other methods of detecting lung cancer. Drs Cata, Kurz, Bauer, and Mascha have reported that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.

Role of sponsors: The sponsor had no involvement in the study design, data acquisition and analysis, or manuscript preparation and review.

AA

arachidonic acid

AUC

area under the curve

FFA

free fatty acid

HETE

hydroxyeicosatetraenoic acid

HODE

hydroxyoctadecadienoic acid

HODE-PC

hydroxyoctadecadienoylphosphatidylcholine

LA

linoleic acid

lysoPAF

lysophospholipid platelet activating factor

PAF

platelet activating factor

Oken MM, Hocking WG, Kvale PA, et al; PLCO Project Team. Screening by chest radiograph and lung cancer mortality: the Prostate, Lung, Colorectal, and Ovarian (PLCO) randomized trial. JAMA. 2011;306(17):1865-1873. [CrossRef] [PubMed]
 
Sox HC. Screening for lung cancer with chest radiographs. JAMA. 2011;306(17):1916-1918. [CrossRef] [PubMed]
 
Bach PB. Reduced lung-cancer mortality with CT screening. N Engl J Med. 2011;365(21):2036. [PubMed]
 
Aberle DR, Adams AM, Berg CD, et al; National Lung Screening Trial Research Team. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365(5):395-409. [CrossRef] [PubMed]
 
Voelker DR, Lee TC, Snyder F. Fatty acid biosynthesis and dietary regulation in pulmonary adenomas. Arch Biochem Biophys. 1976;176(2):753-756. [CrossRef] [PubMed]
 
Kim HA, Kim KJ, Seo KH, Lee HK, Im SY. PTEN/MAPK pathways play a key role in platelet-activating factor-induced experimental pulmonary tumor metastasis. FEBS Lett. 2012;586(24):4296-4302. [CrossRef] [PubMed]
 
Kupert E, Anderson M, Liu Y, et al. Plasma secretory phospholipase A2-IIa as a potential biomarker for lung cancer in patients with solitary pulmonary nodules. BMC Cancer. 2011;11:513. [CrossRef] [PubMed]
 
Zhang M, Xu ZG, Shi Z, et al. Inhibitory effect of celecoxib in lung carcinoma by regulation of cyclooxygenase-2/cytosolic phospholipase A2and peroxisome proliferator-activated receptor gamma. Mol Cell Biochem. 2011;355(1-2):233-240. [CrossRef] [PubMed]
 
Linkous AG, Yazlovitskaya EM, Hallahan DE. Cytosolic phospholipase A2 and lysophospholipids in tumor angiogenesis. J Natl Cancer Inst. 2010;102(18):1398-1412. [CrossRef] [PubMed]
 
Granata F, Frattini A, Loffredo S, et al. Production of vascular endothelial growth factors from human lung macrophages induced by group IIA and group X secreted phospholipases A2. J Immunol. 2010;184(9):5232-5241. [CrossRef] [PubMed]
 
Ho JN, Lee SB, Lee SS, et al. Phospholipase A2 activity of peroxiredoxin 6 promotes invasion and metastasis of lung cancer cells. Mol Cancer Ther. 2010;9(4):825-832. [CrossRef] [PubMed]
 
Weiser-Evans MC, Wang XQ, Amin J, et al. Depletion of cytosolic phospholipase A2 in bone marrow-derived macrophages protects against lung cancer progression and metastasis. Cancer Res. 2009;69(5):1733-1738. [CrossRef] [PubMed]
 
Murakami M, Masuda S, Shimbara S, Ishikawa Y, Ishii T, Kudo I. Cellular distribution, post-translational modification, and tumorigenic potential of human group III secreted phospholipase A(2). J Biol Chem. 2005;280(26):24987-24998. [CrossRef] [PubMed]
 
Murph M, Tanaka T, Pang J, et al. Liquid chromatography mass spectrometry for quantifying plasma lysophospholipids: potential biomarkers for cancer diagnosis. Methods Enzymol. 2007;433:1-25. [PubMed]
 
Gaetano CG, Samadi N, Tomsig JL, Macdonald TL, Lynch KR, Brindley DN. Inhibition of autotaxin production or activity blocks lysophosphatidylcholine-induced migration of human breast cancer and melanoma cells. Mol Carcinog. 2009;48(9):801-809. [CrossRef] [PubMed]
 
Okita M, Gaudette DC, Mills GB, Holub BJ. Elevated levels and altered fatty acid composition of plasma lysophosphatidylcholine (lysoPC) in ovarian cancer patients. Int J Cancer. 1997;71(1):31-34. [CrossRef] [PubMed]
 
Sutphen R, Xu Y, Wilbanks GD, et al. Lysophospholipids are potential biomarkers of ovarian cancer. Cancer Epidemiol Biomarkers Prev. 2004;13(7):1185-1191. [PubMed]
 
Zhou X, Mao J, Ai J, et al. Identification of plasma lipid biomarkers for prostate cancer by lipidomics and bioinformatics. PLoS ONE. 2012;7(11):e48889. [CrossRef] [PubMed]
 
Guo Y, Wang X, Qiu L, et al. Probing gender-specific lipid metabolites and diagnostic biomarkers for lung cancer using Fourier transform ion cyclotron resonance mass spectrometry. Clin Chim Acta. 2012;414:135-141. [CrossRef] [PubMed]
 
Connolly JM, Liu XH, Rose DP. Effects of dietary menhaden oil, soy, and a cyclooxygenase inhibitor on human breast cancer cell growth and metastasis in nude mice. Nutr Cancer. 1997;29(1):48-54. [CrossRef] [PubMed]
 
Rose DP, Connolly JM, Liu XH. Effects of linoleic acid on the growth and metastasis of two human breast cancer cell lines in nude mice and the invasive capacity of these cell lines in vitro. Cancer Res. 1994;54(24):6557-6562. [PubMed]
 
Igal RA. Stearoyl-CoA desaturase-1: a novel key player in the mechanisms of cell proliferation, programmed cell death and transformation to cancer. Carcinogenesis. 2010;31(9):1509-1515. [CrossRef] [PubMed]
 
Hess D, Igal RA. Genistein downregulates de novo lipid synthesis and impairs cell proliferation in human lung cancer cells. Exp Biol Med (Maywood). 2011;236(6):707-713. [CrossRef] [PubMed]
 
Jeon SM, Chandel NS, Hay N. AMPK regulates NADPH homeostasis to promote tumour cell survival during energy stress. Nature. 2012;485(7400):661-665. [CrossRef] [PubMed]
 
Luo DX, Tong DJ, Rajput S, et al. Targeting acetyl-CoA carboxylases: small molecular inhibitors and their therapeutic potential. Recent Patents Anticancer Drug Discov. 2012;7(2):168-184. [CrossRef]
 
Notarnicola M, Messa C, Caruso MG. A significant role of lipogenic enzymes in colorectal cancer. Anticancer Res. 2012;32(7):2585-2590. [PubMed]
 
Kerjaschki D, Bago-Horvath Z, Rudas M, et al. Lipoxygenase mediates invasion of intrametastatic lymphatic vessels and propagates lymph node metastasis of human mammary carcinoma xenografts in mouse. J Clin Invest. 2011;121(5):2000-2012. [CrossRef] [PubMed]
 
Moreno JJ. New aspects of the role of hydroxyeicosatetraenoic acids in cell growth and cancer development. Biochem Pharmacol. 2009;77(1):1-10. [CrossRef] [PubMed]
 
Wang D, DuBois RN. Measurement of eicosanoids in cancer tissues. Methods Enzymol 2007;433:27-50. [PubMed]
 
Paige M, Saprito MS, Bunyan DA, Shim YM. HPLC quantification of 5-hydroxyeicosatetraenoic acid in human lung cancer tissues. Biomed Chromatogr. 2009;23(8):817-821. [CrossRef] [PubMed]
 
Zick SM, Turgeon DK, Vareed SK, et al. Phase II study of the effects of ginger root extract on eicosanoids in colon mucosa in people at normal risk for colorectal cancer. Cancer Prev Res (Phila). 2011;4(11):1929-1937. [CrossRef] [PubMed]
 
Kudriavtsev IA, Miasishcheva NV. Synthesis of arachidonic acid cascade eicosanoids in tumors of various histogenesis in mice [in Russian]. Vestn Ross Akad Med Nauk. 1995;;(4):42-45.
 
Lockwood SF, Penn MS, Hazen SL, Bikádi Z, Zsila F. The effects of oral Cardax (disodium disuccinate astaxanthin) on multiple independent oxidative stress markers in a mouse peritoneal inflammation model: influence on 5-lipoxygenase in vitro and in vivo. Life Sci. 2006;79(2):162-174. [CrossRef] [PubMed]
 
Kelavkar UP, Glasgow W, Olson SJ, Foster BA, Shappell SB. Overexpression of 12/15-lipoxygenase, an ortholog of human 15-lipoxygenase-1, in the prostate tumors of TRAMP mice. Neoplasia. 2004;6(6):821-830. [CrossRef] [PubMed]
 
Avis IM, Jett M, Boyle T, et al. Growth control of lung cancer by interruption of 5-lipoxygenase-mediated growth factor signaling. J Clin Invest. 1996;97(3):806-813. [CrossRef] [PubMed]
 
Ulbricht B, Henny H, Horstmann H, Spring H, Faigle W, Spiess E. Influence of 12(S)-hydroxyeicosatetraenoic acid (12(S)-HETE) on the localization of cathepsin B and cathepsin L in human lung tumor cells. Eur J Cell Biol. 1997;74(3):294-301. [PubMed]
 
Chen YQ, Duniec ZM, Liu B, et al. Endogenous 12(S)-HETE production by tumor cells and its role in metastasis. Cancer Res. 1994;54(6):1574-1579. [PubMed]
 
Honn KV, Nelson KK, Renaud C, Bazaz R, Diglio CA, Timar J. Fatty acid modulation of tumor cell adhesion to microvessel endothelium and experimental metastasis. Prostaglandins. 1992;44(5):413-429. [CrossRef] [PubMed]
 
Grossi IM, Fitzgerald LA, Umbarger LA, et al. Bidirectional control of membrane expression and/or activation of the tumor cell IRGpIIb/IIIa receptor and tumor cell adhesion by lipoxygenase products of arachidonic acid and linoleic acid. Cancer Res. 1989;49(4):1029-1037. [PubMed]
 
Kudriavtsev IA, Miasishcheva NV, Polotskiĭ BE, Machaladze ZO, Davydov MI. Ability of the neoplastic tissue to biosynthesize 12- and 15-hydroxyeicosatetraenoic acids as criterion of metastasizing activity of human lung neoplasms [in Russian]. Biull Eksp Biol Med. 1998;126(9):338-341. [PubMed]
 
Srivastava K, Kundumani-Sridharan V, Zhang B, Bajpai AK, Rao GN. 15(S)-hydroxyeicosatetraenoic acid-induced angiogenesis requires STAT3-dependent expression of VEGF. Cancer Res. 2007;67(9):4328-4336. [CrossRef] [PubMed]
 
Panigrahy D, Greene ER, Pozzi A, Wang DW, Zeldin DC. EET signaling in cancer. Cancer Metastasis Rev. 2011;30(3-4):525-540. [CrossRef] [PubMed]
 
Yuan H, Li MY, Ma LT, et al. 15-Lipoxygenases and its metabolites 15(S)-HETE and 13(S)-HODE in the development of non-small cell lung cancer. Thorax. 2010;65(4):321-326. [CrossRef] [PubMed]
 
Shureiqi I, Jiang W, Zuo X, et al. The 15-lipoxygenase-1 product 13-S-hydroxyoctadecadienoic acid down-regulates PPAR-delta to induce apoptosis in colorectal cancer cells. Proc Natl Acad Sci U S A. 2003;100(17):9968-9973. [CrossRef] [PubMed]
 
Bull AW, Branting C, Bronstein JC, Blackburn ML, Rafter J. Increases in 13-hydroxyoctadecadienoic acid dehydrogenase activity during differentiation of cultured cells. Carcinogenesis. 1993;14(11):2239-2243. [CrossRef] [PubMed]
 
Buchanan MR, Bertomeu MC, Haas TA, Gallo S, Eltringham-Smith L. Endothelial cell 13-HODE synthesis and tumor cell endothelial cell adhesion. Adv Prostaglandin Thromboxane Leukot Res. 1991;21B:909-912. [PubMed]
 
Nickkho-Amiry M, McVey R, Holland C. Peroxisome proliferator-activated receptors modulate proliferation and angiogenesis in human endometrial carcinoma. Mol Cancer Res. 2012;10(3):441-453. [CrossRef] [PubMed]
 
Li H, Sorenson AL, Poczobutt J, et al. Activation of PPARγ in myeloid cells promotes lung cancer progression and metastasis. PLoS ONE. 2011;6(12):e28133. [CrossRef] [PubMed]
 
van Eijsden M, van der Wal MF, Hornstra G, Bonsel GJ. Can whole-blood samples be stored over 24 hours without compromising stability of C-reactive protein, retinol, ferritin, folic acid, and fatty acids in epidemiologic research? Clin Chem. 2005;51(1):230-232. [CrossRef] [PubMed]
 
Salo MK, Gey F, Nikkari T. Stability of plasma fatty acids at -20 degrees C and its relationship to antioxidants. Int J Vitam Nutr Res. 1986;56(3):231-239. [PubMed]
 
Rogiers V. Stability of the long chain non-esterified fatty acid pattern in plasma and blood during different storage conditions. Clin Chim Acta. 1978;84(1-2):49-54. [CrossRef] [PubMed]
 
Hirsch EZ, Slivka S, Gibbons AP. Stability of fatty acids in hyperlipoproteinemic plasma during long-term storage. Clin Chem. 1976;22(4):445-448. [PubMed]
 
Matthan NR, Ip B, Resteghini N, Ausman LM, Lichtenstein AH. Long-term fatty acid stability in human serum cholesteryl ester, triglyceride, and phospholipid fractions. J Lipid Res. 2010;51(9):2826-2832. [CrossRef] [PubMed]
 

Figures

Figure Jump LinkFigure 1  The role of LPs in tumor development. LP binds its cognate GPCR and stimulates cell proliferation and migration, which may initiate cancer cell development and metastasis. GPCR = G-protein-coupled receptor; LP = lysophospholipid.Grahic Jump Location
Figure Jump LinkFigure 2  Box plots for predictors of adenocarcinoma in patients with and without cancer. Displayed predictors needed to have area under the receiver operating characteristics curve significantly > 0.50 and both sensitivity and specificity > 0.70 at a “best” cutpoint, which maximizes both parameters. Top and bottom of box are first and third quartiles, respectively. Solid line in side box is median, diamond represents the mean. Whiskers extend to the minimum of 1.5 interquartile ranges or the range of the data. AA = arachidonic acid; CA = cancer; HETE = hydroxyeicosatetraenoic acid; LA = linoleic acid.Grahic Jump Location
Figure Jump LinkFigure 3  Receiver operating characteristic curves with area under the curve (AUC) and SE in parentheses for patients with adenocarcinoma. Displayed are the three variables that had AUC significantly > 0.50 and both sensitivity and specificity estimated as > 0.70 (see Table 5). See Figure 2 legend for expansion of abbreviations.Grahic Jump Location

Tables

Table Graphic Jump Location
TABLE 1  ] Matched Patients With Adenocarcinoma and Control Patients: Baseline Characteristics

Results presented as No. (%) unless otherwise noted. DM = diabetes mellitus; F = Fisher exact test; STD = standardized difference; W = Wilcoxon rank sum test.

a 

Pearson χ2 test, unless otherwise noted.

b 

Difference in means or proportions/pooled SD.

Table Graphic Jump Location
TABLE 2  ] Matched Patients With Squamous Cell Carcinoma and Control Patients: Baseline Characteristics

Results presented as No. (%) unless otherwise noted. See Table 1 legend for expansion of abbreviations.

a 

Pearson χ2 test, unless otherwise noted.

b 

Difference in means or proportions/pooled SD.

Table Graphic Jump Location
TABLE 3  ] Adenocarcinoma: Comparing Patients With Cancer and Control Patients on Fatty Acids and Phospholipids

AA = arachidonic acid; AzPAF = azelaoyl platelet activating factor; AzPC = azelaoyl phosphatidylcholine; HETE = hydroxyeicosatetraenoic acid; HODE = hydroxyoctadecadienoic acid; HODE-PC = hydroxyoctadecadienoylphosphatidylcholine; HpODE-PC =hydroperoxyoctadecadienoyl phosphatidylcholine; LA = linoleic acid; LPC = lyso-phosphatidylcholine; lysoPAF = lysophospholipid platelet activating factor; Q1 = first quartile = 25th percentile; Q3 = third quartile = 75th percentile. See Table 1 legend for expansion of other abbreviation.

a 

Estimated ratio of geometric means, calculated as exponentiated difference in means of the log-transformed data; 99.7% CI maintains type 1 error at 5% across 15 variables.

b 

Two-tailed Wilcoxon rank sum test.

c 

Significant if P < .05/15 = .0033.

Table Graphic Jump Location
TABLE 4  ] Squamous Cell: Comparing Patients With Cancer and Control Patients on Fatty Acid and Phospholipids

See Table 1 and 3 legends for expansion of abbreviations.

a 

Estimated ratio of geometric means, calculated as exponentiated difference in means of the log-transformed data; 99.7% CI maintains type 1 error at 5% across 15 variables.

b 

Two-tailed Wilcoxon rank sum test.

c 

Difference in means or proportions/pooled SD.

Table Graphic Jump Location
TABLE 5  ] Adenocarcinoma: Diagnostic Accuracy of Each FFA Metabolite Predicting Lung Cancer

N = 37 patients with cancer, 111 control subjects. Diagnostic accuracy parameters (sensitivity, specificity, and so forth) correspond to the given threshold value. CI estimated using bootstrap resampling (percentile method) with 5,000 resamples. Accuracy: ratio of sum of true positives, true negatives, false positives, false negatives to total sample size. Data values represent peak raw biomarker values divided by the corresponding internal standard for each sample. N = 3 biomarkers met the criteria of having sensitivity and specificity at the best cutpoint > 0.70 (AA, LA, HETE15). CI: 99.7% CIs adjusted for assessing multiple biomarkers using Bonferroni correction; overall α = 0.05. FFA metabolite values are metabolite peak/internal standard peak. AUC = area under the curve; FFA = free fatty acid; NPV = negative predictive value; PPV = positive predictive value. See Table 3 legend for expansion of other abbreviations.

a 

All P values < .001 except for AzPAF (P = .027), HpODE-PC (P = .040).

b 

Jointly maximizes sensitivity and specificity.

c 

Estimate using Bayes’ theorem assuming a true prevalence of 0.10.

Table Graphic Jump Location
TABLE 6  ] Squamous Cell: Diagnostic Accuracy of Each FFA Metabolite Predicting Lung Cancer

N = 18 patients with cancer, 54 control subjects. FFA metabolite values are metabolite peak/internal standard peak. Diagnostic accuracy parameters (sensitivity, specificity, etc.) correspond to the given threshold value. CI estimated using bootstrap resampling (percentile method) with 5,000 resamples. Accuracy: ratio of sum of true positives, true negatives, false positives, false negatives to total sample size. Data values represent peak raw biomarker values divided by the corresponding internal standard for each sample. No biomarkers met the criteria of having sensitivity and specificity at the best cutpoint > 0.70. CI: 99.67% CIs adjusted for assessing multiple biomarkers using Bonferroni correction; overall α = 0.05 CI: 99.7% CIs adjusted for assessing multiple biomarkers using Bonferroni correction; overall α = 0.05. See Table 3 and 5 legends for expansion of abbreviations.

a 

Jointly maximizes sensitivity and specificity.

b 

Estimate using Bayes’ theorem assuming a true prevalence of 0.10. AUC P value: testing H0: AUC = 0.5 (1-tailed).

References

Oken MM, Hocking WG, Kvale PA, et al; PLCO Project Team. Screening by chest radiograph and lung cancer mortality: the Prostate, Lung, Colorectal, and Ovarian (PLCO) randomized trial. JAMA. 2011;306(17):1865-1873. [CrossRef] [PubMed]
 
Sox HC. Screening for lung cancer with chest radiographs. JAMA. 2011;306(17):1916-1918. [CrossRef] [PubMed]
 
Bach PB. Reduced lung-cancer mortality with CT screening. N Engl J Med. 2011;365(21):2036. [PubMed]
 
Aberle DR, Adams AM, Berg CD, et al; National Lung Screening Trial Research Team. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365(5):395-409. [CrossRef] [PubMed]
 
Voelker DR, Lee TC, Snyder F. Fatty acid biosynthesis and dietary regulation in pulmonary adenomas. Arch Biochem Biophys. 1976;176(2):753-756. [CrossRef] [PubMed]
 
Kim HA, Kim KJ, Seo KH, Lee HK, Im SY. PTEN/MAPK pathways play a key role in platelet-activating factor-induced experimental pulmonary tumor metastasis. FEBS Lett. 2012;586(24):4296-4302. [CrossRef] [PubMed]
 
Kupert E, Anderson M, Liu Y, et al. Plasma secretory phospholipase A2-IIa as a potential biomarker for lung cancer in patients with solitary pulmonary nodules. BMC Cancer. 2011;11:513. [CrossRef] [PubMed]
 
Zhang M, Xu ZG, Shi Z, et al. Inhibitory effect of celecoxib in lung carcinoma by regulation of cyclooxygenase-2/cytosolic phospholipase A2and peroxisome proliferator-activated receptor gamma. Mol Cell Biochem. 2011;355(1-2):233-240. [CrossRef] [PubMed]
 
Linkous AG, Yazlovitskaya EM, Hallahan DE. Cytosolic phospholipase A2 and lysophospholipids in tumor angiogenesis. J Natl Cancer Inst. 2010;102(18):1398-1412. [CrossRef] [PubMed]
 
Granata F, Frattini A, Loffredo S, et al. Production of vascular endothelial growth factors from human lung macrophages induced by group IIA and group X secreted phospholipases A2. J Immunol. 2010;184(9):5232-5241. [CrossRef] [PubMed]
 
Ho JN, Lee SB, Lee SS, et al. Phospholipase A2 activity of peroxiredoxin 6 promotes invasion and metastasis of lung cancer cells. Mol Cancer Ther. 2010;9(4):825-832. [CrossRef] [PubMed]
 
Weiser-Evans MC, Wang XQ, Amin J, et al. Depletion of cytosolic phospholipase A2 in bone marrow-derived macrophages protects against lung cancer progression and metastasis. Cancer Res. 2009;69(5):1733-1738. [CrossRef] [PubMed]
 
Murakami M, Masuda S, Shimbara S, Ishikawa Y, Ishii T, Kudo I. Cellular distribution, post-translational modification, and tumorigenic potential of human group III secreted phospholipase A(2). J Biol Chem. 2005;280(26):24987-24998. [CrossRef] [PubMed]
 
Murph M, Tanaka T, Pang J, et al. Liquid chromatography mass spectrometry for quantifying plasma lysophospholipids: potential biomarkers for cancer diagnosis. Methods Enzymol. 2007;433:1-25. [PubMed]
 
Gaetano CG, Samadi N, Tomsig JL, Macdonald TL, Lynch KR, Brindley DN. Inhibition of autotaxin production or activity blocks lysophosphatidylcholine-induced migration of human breast cancer and melanoma cells. Mol Carcinog. 2009;48(9):801-809. [CrossRef] [PubMed]
 
Okita M, Gaudette DC, Mills GB, Holub BJ. Elevated levels and altered fatty acid composition of plasma lysophosphatidylcholine (lysoPC) in ovarian cancer patients. Int J Cancer. 1997;71(1):31-34. [CrossRef] [PubMed]
 
Sutphen R, Xu Y, Wilbanks GD, et al. Lysophospholipids are potential biomarkers of ovarian cancer. Cancer Epidemiol Biomarkers Prev. 2004;13(7):1185-1191. [PubMed]
 
Zhou X, Mao J, Ai J, et al. Identification of plasma lipid biomarkers for prostate cancer by lipidomics and bioinformatics. PLoS ONE. 2012;7(11):e48889. [CrossRef] [PubMed]
 
Guo Y, Wang X, Qiu L, et al. Probing gender-specific lipid metabolites and diagnostic biomarkers for lung cancer using Fourier transform ion cyclotron resonance mass spectrometry. Clin Chim Acta. 2012;414:135-141. [CrossRef] [PubMed]
 
Connolly JM, Liu XH, Rose DP. Effects of dietary menhaden oil, soy, and a cyclooxygenase inhibitor on human breast cancer cell growth and metastasis in nude mice. Nutr Cancer. 1997;29(1):48-54. [CrossRef] [PubMed]
 
Rose DP, Connolly JM, Liu XH. Effects of linoleic acid on the growth and metastasis of two human breast cancer cell lines in nude mice and the invasive capacity of these cell lines in vitro. Cancer Res. 1994;54(24):6557-6562. [PubMed]
 
Igal RA. Stearoyl-CoA desaturase-1: a novel key player in the mechanisms of cell proliferation, programmed cell death and transformation to cancer. Carcinogenesis. 2010;31(9):1509-1515. [CrossRef] [PubMed]
 
Hess D, Igal RA. Genistein downregulates de novo lipid synthesis and impairs cell proliferation in human lung cancer cells. Exp Biol Med (Maywood). 2011;236(6):707-713. [CrossRef] [PubMed]
 
Jeon SM, Chandel NS, Hay N. AMPK regulates NADPH homeostasis to promote tumour cell survival during energy stress. Nature. 2012;485(7400):661-665. [CrossRef] [PubMed]
 
Luo DX, Tong DJ, Rajput S, et al. Targeting acetyl-CoA carboxylases: small molecular inhibitors and their therapeutic potential. Recent Patents Anticancer Drug Discov. 2012;7(2):168-184. [CrossRef]
 
Notarnicola M, Messa C, Caruso MG. A significant role of lipogenic enzymes in colorectal cancer. Anticancer Res. 2012;32(7):2585-2590. [PubMed]
 
Kerjaschki D, Bago-Horvath Z, Rudas M, et al. Lipoxygenase mediates invasion of intrametastatic lymphatic vessels and propagates lymph node metastasis of human mammary carcinoma xenografts in mouse. J Clin Invest. 2011;121(5):2000-2012. [CrossRef] [PubMed]
 
Moreno JJ. New aspects of the role of hydroxyeicosatetraenoic acids in cell growth and cancer development. Biochem Pharmacol. 2009;77(1):1-10. [CrossRef] [PubMed]
 
Wang D, DuBois RN. Measurement of eicosanoids in cancer tissues. Methods Enzymol 2007;433:27-50. [PubMed]
 
Paige M, Saprito MS, Bunyan DA, Shim YM. HPLC quantification of 5-hydroxyeicosatetraenoic acid in human lung cancer tissues. Biomed Chromatogr. 2009;23(8):817-821. [CrossRef] [PubMed]
 
Zick SM, Turgeon DK, Vareed SK, et al. Phase II study of the effects of ginger root extract on eicosanoids in colon mucosa in people at normal risk for colorectal cancer. Cancer Prev Res (Phila). 2011;4(11):1929-1937. [CrossRef] [PubMed]
 
Kudriavtsev IA, Miasishcheva NV. Synthesis of arachidonic acid cascade eicosanoids in tumors of various histogenesis in mice [in Russian]. Vestn Ross Akad Med Nauk. 1995;;(4):42-45.
 
Lockwood SF, Penn MS, Hazen SL, Bikádi Z, Zsila F. The effects of oral Cardax (disodium disuccinate astaxanthin) on multiple independent oxidative stress markers in a mouse peritoneal inflammation model: influence on 5-lipoxygenase in vitro and in vivo. Life Sci. 2006;79(2):162-174. [CrossRef] [PubMed]
 
Kelavkar UP, Glasgow W, Olson SJ, Foster BA, Shappell SB. Overexpression of 12/15-lipoxygenase, an ortholog of human 15-lipoxygenase-1, in the prostate tumors of TRAMP mice. Neoplasia. 2004;6(6):821-830. [CrossRef] [PubMed]
 
Avis IM, Jett M, Boyle T, et al. Growth control of lung cancer by interruption of 5-lipoxygenase-mediated growth factor signaling. J Clin Invest. 1996;97(3):806-813. [CrossRef] [PubMed]
 
Ulbricht B, Henny H, Horstmann H, Spring H, Faigle W, Spiess E. Influence of 12(S)-hydroxyeicosatetraenoic acid (12(S)-HETE) on the localization of cathepsin B and cathepsin L in human lung tumor cells. Eur J Cell Biol. 1997;74(3):294-301. [PubMed]
 
Chen YQ, Duniec ZM, Liu B, et al. Endogenous 12(S)-HETE production by tumor cells and its role in metastasis. Cancer Res. 1994;54(6):1574-1579. [PubMed]
 
Honn KV, Nelson KK, Renaud C, Bazaz R, Diglio CA, Timar J. Fatty acid modulation of tumor cell adhesion to microvessel endothelium and experimental metastasis. Prostaglandins. 1992;44(5):413-429. [CrossRef] [PubMed]
 
Grossi IM, Fitzgerald LA, Umbarger LA, et al. Bidirectional control of membrane expression and/or activation of the tumor cell IRGpIIb/IIIa receptor and tumor cell adhesion by lipoxygenase products of arachidonic acid and linoleic acid. Cancer Res. 1989;49(4):1029-1037. [PubMed]
 
Kudriavtsev IA, Miasishcheva NV, Polotskiĭ BE, Machaladze ZO, Davydov MI. Ability of the neoplastic tissue to biosynthesize 12- and 15-hydroxyeicosatetraenoic acids as criterion of metastasizing activity of human lung neoplasms [in Russian]. Biull Eksp Biol Med. 1998;126(9):338-341. [PubMed]
 
Srivastava K, Kundumani-Sridharan V, Zhang B, Bajpai AK, Rao GN. 15(S)-hydroxyeicosatetraenoic acid-induced angiogenesis requires STAT3-dependent expression of VEGF. Cancer Res. 2007;67(9):4328-4336. [CrossRef] [PubMed]
 
Panigrahy D, Greene ER, Pozzi A, Wang DW, Zeldin DC. EET signaling in cancer. Cancer Metastasis Rev. 2011;30(3-4):525-540. [CrossRef] [PubMed]
 
Yuan H, Li MY, Ma LT, et al. 15-Lipoxygenases and its metabolites 15(S)-HETE and 13(S)-HODE in the development of non-small cell lung cancer. Thorax. 2010;65(4):321-326. [CrossRef] [PubMed]
 
Shureiqi I, Jiang W, Zuo X, et al. The 15-lipoxygenase-1 product 13-S-hydroxyoctadecadienoic acid down-regulates PPAR-delta to induce apoptosis in colorectal cancer cells. Proc Natl Acad Sci U S A. 2003;100(17):9968-9973. [CrossRef] [PubMed]
 
Bull AW, Branting C, Bronstein JC, Blackburn ML, Rafter J. Increases in 13-hydroxyoctadecadienoic acid dehydrogenase activity during differentiation of cultured cells. Carcinogenesis. 1993;14(11):2239-2243. [CrossRef] [PubMed]
 
Buchanan MR, Bertomeu MC, Haas TA, Gallo S, Eltringham-Smith L. Endothelial cell 13-HODE synthesis and tumor cell endothelial cell adhesion. Adv Prostaglandin Thromboxane Leukot Res. 1991;21B:909-912. [PubMed]
 
Nickkho-Amiry M, McVey R, Holland C. Peroxisome proliferator-activated receptors modulate proliferation and angiogenesis in human endometrial carcinoma. Mol Cancer Res. 2012;10(3):441-453. [CrossRef] [PubMed]
 
Li H, Sorenson AL, Poczobutt J, et al. Activation of PPARγ in myeloid cells promotes lung cancer progression and metastasis. PLoS ONE. 2011;6(12):e28133. [CrossRef] [PubMed]
 
van Eijsden M, van der Wal MF, Hornstra G, Bonsel GJ. Can whole-blood samples be stored over 24 hours without compromising stability of C-reactive protein, retinol, ferritin, folic acid, and fatty acids in epidemiologic research? Clin Chem. 2005;51(1):230-232. [CrossRef] [PubMed]
 
Salo MK, Gey F, Nikkari T. Stability of plasma fatty acids at -20 degrees C and its relationship to antioxidants. Int J Vitam Nutr Res. 1986;56(3):231-239. [PubMed]
 
Rogiers V. Stability of the long chain non-esterified fatty acid pattern in plasma and blood during different storage conditions. Clin Chim Acta. 1978;84(1-2):49-54. [CrossRef] [PubMed]
 
Hirsch EZ, Slivka S, Gibbons AP. Stability of fatty acids in hyperlipoproteinemic plasma during long-term storage. Clin Chem. 1976;22(4):445-448. [PubMed]
 
Matthan NR, Ip B, Resteghini N, Ausman LM, Lichtenstein AH. Long-term fatty acid stability in human serum cholesteryl ester, triglyceride, and phospholipid fractions. J Lipid Res. 2010;51(9):2826-2832. [CrossRef] [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543