With increased sensitivity of imaging technology, such as low dose spiral computed tomography (LDCT), lung cancer can be detected at an earlier stage [2
]. Recently, lung cancer screening with CT has been shown to detect lung tumors earlier and decrease lung cancer mortality [30
]. One of the problems with the use of CT screening is the number of false positive with up to 25% of the subjects in the lung cancer screening trials having nodules, most of which are benign. Given the epidemic of histoplasmosis in the Ohio valley, we found that among 132 people enrolled in a pilot lung cancer screening project, 61% of subjects had nodules, most of which were benign [5
]. If lung cancer screening with CT becomes standard, there will be a large number of nodules that will need to be evaluated. Monitoring these nodules usually requires repeated CT scans at 3, 6, 9, 12, 24 months based on the sizes of SPNs [6
]. An increase in the size of a SPN is currently the parameter used to predict the presence of lung cancer. This repeated CT scan results in radiation exposure as well as anxiety. Patients may also undergo invasive procedures for diagnosis such as bronchoscopy, CT-guided fine needle aspiration, or thoracoscopic resection. It will become critical to have more effective ways to determine which nodules are most likely malignant.
We recently reported that there is an elevated level of plasma sPLA2-IIa in prostate cancer patients [17
]. The current study showed that lung cancer tissues overexpress sPLA2-IIa and that patients with lung cancer have elevated levels of plasma sPLA2-IIa. The expression of the sPLA2-IIa gene is not tissue-specific or cancer-specific [17
]. It was reported that the levels of plasma sPLA2-IIa are elevated with bacterial and viral infection or IL-2 infusion [31
] and in coronary heart disease [33
]. However, a recent clinical study including more than 500 patients showed that there is no significant alteration in plasma level of sPLA2-IIa protein among patients with coronary artery disease relative to healthy controls [34
]. In consideration of these complications, we included some patients with inflammatory nodules of the lung as controls. We found an increased basal level of plasma sPLA2-IIa in these controls relative to those in healthy donors, which may be a result of chronic inflammation due to fungal infection and pneumonia. This active localized inflammation can occasionally lead to a moderate increased level of plasma sPLA2-IIa. Nevertheless, a significant high level of plasma sPLA2-IIa as a result of lung cancer can predict more than 48% of early stage lung cancers and up to 67% of T2 stage lung tumors with 86% specificity. sPLA2-IIa blood test has potential to help with the decision algorithm and determine the timing of subsequent CT scans and potential biopsy of concerning nodules.
One concern for using sPLA2-IIa as a lung cancer biomarker is that it lacks cancer specificity. This is one of the common features shared by all biomarkers currently in clinical use or under development and is simply due to the fact that genes contributing to cancer development overlap with those for other metabolic diseases, such as diabetes, obesity, and inflammation. In addition, the sensitivity of these plasma biomarkers is usually not very high, especially for early stages of cancer, which is partially due to the heterogeneous nature of cancer. Only a few plasma biomarkers are currently used to screen and diagnose cancers in clinical practice, including PSA, α-fetoprotein, CA19-9, CA125, and CEA [9
]. None of these plasma biomarkers are cancer-specific [11
]. For example, PSA is tissue-specific but not cancer-specific and has only 21% sensitivity with 4 ng/ml as the cutoff value [36
]. Plasma PSA are elevated in benign diseases, such as in 30 ~ 50% BPH patients, which leads to low specificity [37
]. On the other hand, many prostate cancers do not lead to a high PSA and fail to be detected in PSA screening, which also contributes to low sensitivity [37
]. α-fetoprotein (AFP) for hepatocellular carcinoma has 65% sensitivity and 89% specificity, while the addition of VEGF and fucosidase (AFU) tests can increase the sensitivity up to 100% [42
]. Furthermore, it is also noted that sensitivity of plasma biomarkers, such as CA125, is increased with advanced cancer stages, which is also consistent with our observation of plasma sPLA2-IIa for lung cancer [44
A few serum biomarkers for lung cancer are currently under investigation, including Cyfra 21.1 and CEA for NSCLC and neuron-specific enolase (NSE) for SCLC [12
]. Our data showed that none of these lung cancer biomarkers, based on the cutoff value reported, has high sensitivity to predict an early stage lung cancer [45
]. However, the combined test of sPLA2-IIa, Cyfra 21.1 and CEA increases the power for lung cancer prediction relative to sPLA2-IIa test alone. We did not examine NSE, since there are only three SCLC specimens in the BNLCC cohort. Although Cyfra 21.1 is preferentially for squamous cell carcinoma, we did not find that Cyfra 21.1 shows a high sensitivity for squamous cell carcinoma. Given the heterogeneous nature of cancer and heterogeneous gene overexpression from one cancer to another, a combined blood test including a few plasma biomarkers is essential to increase the sensitivity to predict an early stage lung cancer.
In summary, we demonstrated that lung cancer tissues overexpress sPLA2-IIa and there is an elevated level of plasma sPLA2-IIa in lung cancer patients. Multiple lines of evidence support the notion that plasma sPLA2-IIa may represent a novel biomarker for lung cancer: 1) sPLA2-IIa is overexpressed in all squamous cell carcinoma, adenocarcinoma, and bronchioalveolar carcinoma examined and in majority of small cell carcinoma; 2) High levels of plasma sPLA2-IIa predict approximately 48% of early stage lung cancers and up to 67% of T2 stage lung cancers relative to patients with benign SPN; 3) High levels of plasma sPLA2-IIa are associated with advanced lung cancer stages and a decreased overall cancer survival; 4) the combined sPLA2-IIa, Cyfra 21.1 and CEA tests increase the power for lung cancer prediction relative to sPLA2-IIa test alone; 5) The combined lung cancer blood tests have potential to help with the decision algorithm of lung cancer diagnosis and reduce the number of CT scan and radiation exposure.