Circulating microparticles are prognostic biomarkers in advanced non-small cell lung cancer patients

We investigated whether circulating microparticles (MPs) could serve as prognostic biomarkers in non-small cell lung cancer (NSCLC) patients. We enrolled 25 control subjects and 136 NSCLC patients categorized into disease-progression (DP, n=42) and disease-control (DC, n=94) groups. Flow cytometric analysis showed that levels of four types of circulating microparticles (EDAc-MPs, EDAp-MPs, PDAc-MPs and PDAp-MPs) were higher in the study patients than the control subjects (P < 0.04). DP patients showed poor initially performance status and more non-adenocarcinomas than DC patients. DC patients showed more EGFR mutations and poorer performance to targeted therapy than DP patients (P < 0.01). Three months after therapy, the levels of all four types of circulating MPs were lower in DC than DP patients (P < 0.02), and were comparable to the levels in control subjects. In addition, the levels of circulating MPs after 3 months accurately predicted one-year prognostic outcomes (P < 0.05). This study showed that circulating MPs are valuable prognostic biomarkers in advanced NSCLC patients.

Circulating MPs are also associated with a wide range of diseases including LC [28,[32][33][34][35][36][37][38][39]. Circulating endothelial-derived activated MPs (EDAc-MPs) were useful in predicting 1-year morality in advanced stage NSCLC patients [20]. However, since majority of the patients had received palliative treatment prior to enrolment in our previous study, the PDAp-MPs (platelet-derived apoptotic MPs), PDAc-MPs (platelet-derived activated MPs), and EDAp-MPs (endothelial-derived apoptotic MPs) were not prognostic [20]. Therefore, we conducted this prospective study by measuring the circulating levels of MPs in advanced stage NSCLC patients to analyze their prognostic outcomes in advanced stage NSCLC patients.

Baseline circulating levels of four types microparticles in study subjects
The circulating levels of the PDAc-MPs, PDAp-MPs, EDAc-MPs and EDAp-MPs were significantly higher in advanced NSCLC patients compared to control subjects. This suggested that the circulating MPs are useful diagnostic biomarkers for advanced NSCLC patients (Table 1 and Figure 1).

Baseline characteristics of study patients
The disease control (DC) and disease progression (DP) groups had similar parameters like gender, weight, height, surface area, mass index and incidence of smoking status. The serum levels of total cholesterol, sugar, glutamic oxaloacetic transaminase (GOT) and glutamic pyruvic transaminase (GPT) were also similar between these two groups. Furthermore, the red blood cell (RBC), white blood cell (WBC) and platelet counts as well as circulating levels of carcinoembryonic antigen (CEA) were also similar between the 2 groups. Furthermore, comorbidities like hypertension, diabetes mellitus, coronary artery disease and chronic obstructive lung disease were comparable between DC and DP groups ( Table 2).

Lung cancer associated parameters in the study patients
The cell types of lung cancer (adenocarcinoma or non-adenocarcinoma) were similar between the DC and the DP groups. However, higher epidermal growth factor receptor (EGFR) mutation levels were observed in the DC patients compared to DP group. The incidence of metastasis and the metastatic sites were similar for the two groups of the patients. Also, the two groups showed no differences in stages IIIB or IV. However, the DP patients' performance status was poorer than the DC group, and the DC group of patients was easily met with target therapy (Table 3). Table 4 shows the changes in circulating levels of PDAc-MPs, PDAp-MPs, EDAc-MPs, EDAp-MPs between the DC and DP patients prior to and at the end of 1 st and 3 rd months after pharmacological intervention. The circulating levels of the four types of MPs were similar between the DC and DP groups prior to and at the end of the first month after pharmacological therapy ( Figure 2). However, at the end of the third month after pharmacological therapy, all the four types of MPs were significantly lower in the DC group than in the DP group ( Figure 2). This was corroborated by positive ΔEDAp-MP, ΔEDAc-MP, ΔPDAp-MP and ΔPDAc-MP values associated with DP. These findings demonstrated that the pharmacological therapy lowered the circulating MPs in the DC group (Table 4 and Figure 2). Table 5 compares changes in circulating levels of PDAc-MPs, PDAp-MPs, EDAc-MPs, EDAp-MPs in the patients that received chemotherapy or targeted therapy. The lung cancer patients with EGFR mutant type were treated with EGFR TKI agents and those with EGFR wildtype subgroup were treated with chemotherapy based on guideline recommendations. The flow cytometric analysis showed that the circulating levels of the four types of MPs were similar between the chemotherapy and targeted therapy patients prior to and at the end of the first month after pharmacological therapy. However, at the end of the third month after pharmacological therapy, all the four types of MPs were significantly lower in patients that www.impactjournals.com/oncotarget

Circulating levels of microparticles in one-year survivors and non-survivors
We observed that the four types of circulating MPs did not differ between 1-year survivors and nonsurvivors prior to and at one month time intervals after pharmacological intervention. However, at the end of the third month after pharmacological intervention, except for EDAc-MPs, the other three types of MPs were significantly higher in the one-year non-survivors than in one-year survivors. Additionally, PDAc-MPs and EDAc-MPs levels were significantly higher in 1-year non-survivors than in 1-year survivors. These findings suggest that the circulating MPs can serve as 1-year prognostic predictors in advanced stage NSCLC patients. (Table 6 and Figure 3)

Receiver operating characteristic (ROC) plot of circulating levels of MPs and CEA in NSCLC patients with progressive and non-progressive disease
Next, we examined if the circulating levels of the four types of MPs and CEA could predict progressive Data are expressed as mean ± SD or % (n). disease in NSCLC patients. As shown in Figure 4 and Table 7, ROC curve analysis showed that the circulating levels of EDAp-MPs, EDAc-MPs, PDAp-MPs and PDAc-MPs were greater than 10468.5 counts/ ml, 3557 counts/ml, 15055 counts/ml and 62700.5 counts/ml, respectively at the 3 rd month after therapy. These data showed that the 4 types of MPs had high sensitivity and specificity and were good prognostic predictors for advanced stage NSCLC. Further, multivariate logistic regression analysis showed that the EGFR mutant, levels of EDAc-MPs at 3 rd month after therapy and the ΔPDAp-MPs and ΔPDAc-MPs were independent prognostic predictors in NSCLC patients (Table 8).

DISCUSSION
In this study, we determined if circulating levels of MPs could predict clinical outcomes in advanced stage NSCLC patients. We observed that circulating levels of the four types of MPs that we analyzed were higher in advanced stage NSCLC patients. The levels of EDAc-MPs, EDAp-MPs, PDAc-MPs, PDAp-MPs did not differ between DC and DP prior to therapeutic intervention. However, by the end of the third month, all the four biomarkers were significantly lower in the DC group compared with the DP group and similar to control subjects. Additionally, the positive net changes of four types of microparticles (i.e., ΔEDAp-MPs, ΔEDAc-MPs,  Data expressed as mean ± SD. PDAc-MPs = platelet-derived activated microparticles; PDAp-MPs = platelet-derived apoptotic MPs; EDAc-MPs = endothelial-derived activated MPs; EDAp-MPs = endothelial-derived apoptotic MPs. * indicated the blood sampling was performed prior to any treatment. † indicated the blood sampling was performed at the end of 1 st month after pharmacological intervention. ‡ indicated the blood sample was performed at the end of the 3 rd month after pharmacological intervention. ΔEDAp-MPs: EDApMPs levels that three months after treatment minus initial level ΔEDAc-MPs: EDAcMPs levels that three months after treatment minus initial level ΔPDAp-MPs: PDApMPs levels that three months after treatment minus initial level ΔPDAc-MPs: PDAcMPs levels that three months after treatment minus initial level ΔPDAp-MPs and ΔPDAc-MPs) between 3 rd month and baseline were strongly associated with DP. Furthermore, ROC curve identified that the absolute values at three months and the relative changes of MP values over time (i.e., ΔEDAp-MPs, ΔEDAc-MPs, ΔPDAp-MPs and ΔPDAc-MPs) had notably higher sensitivity and specificity than that of the CEA level for predicting the prognostic outcomes. Moreover, multivariate logistic regression analysis exhibited that the EGFR mutant, EDAc-MPs level at 3 rd month and the net change between baseline and 3 rd month (i.e., ΔPDAp-MPs and ΔPDAc-MPs) were also independently predictive of DP in NSCLC patients. Accordingly, these suggested that these four types of circulating MPs may be useful biomarkers for predicting prognostic outcomes in LC patients. This study also showed that the circulating levels EDAc-MPs, EDAp-MPs, PDAc-MPs were higher in study patients prior to receiving treatment than in control subjects and were consistent with previous studies [18][19][20]. Most importantly, the circulating levels of the four types of MPs in the DC group were significantly lower and comparable to the control subjects at end of the third month in the DC group than in the DP group. Furthermore, the circulating levels of MPs independently predicted the one-year prognostic clinical outcome in the advanced stage NSCLC patients. Thus, our findings highlight that serial measurement of circulating MPs can predict therapeutic response and prognostic outcomes in advanced NSCLC patients, especially when considering the cost and effectiveness of target therapy.
An association between aberrant EGFR mutation activity and better prognostic outcomes has been reported previously [40] [18][19][20]. Additionally, poor performance status and poorer prognostic outcomes were  Data expressed as mean ± SD. PDAc-MPs = platelet-derived activated microparticles; PDAp-MPs = platelet-derived apoptotic MPs; EDAc-MPs = endothelial-derived activated MPs; EDAp-MPs = endothelial-derived apoptotic MPs. * indicated the blood sampling was performed prior to any treatment. † indicated the blood sampling was performed at the end of 1 st month after pharmacological intervention. ‡ indicated the blood sample was performed at the end of the 3 rd month after pharmacological intervention.    NSCLC = non small cell lung cancer; AUC = area under the curve; CI = confidence interval; NSCLC = non small cell lung cancer; PDAc-MPs = platelet-derived activated microparticles; PDAp-MPs = platelet-derived apoptotic microparticles; EDAc-MPs = endothelial derived activated microparticles; EDAp-MPs = endothelial-derived apoptotic microparticles. ΔEDAp-MPs: EDApMPs levels that three months after treatment minus initial level ΔEDAc-MPs: EDAcMPs levels that three months after treatment minus initial level ΔPDAp-MPs: PDApMPs levels that three months after treatment minus initial level ΔPDAc-MPs: PDAcMPs levels that three months after treatment minus initial level  associated according to previous studies [41] [18][19][20]. We demonstrated that the poor performance status was significantly higher in the DP group, whereas EGFR mutations were significantly lower in the DP group. This study has limitations. First, the sample size of this cohort study was relatively small. Therefore, a new clinical trial with larger sample size is needed to validate that circulating MP levels can predict long-term clinical outcomes in advanced NSCLC patients. Second, this study measured only advanced NSCLC patients. Therefore, the usefulness of these four biomarkers in small cell lung cancer or other types of cancer patients is not known.
In conclusion, we demonstrated that serial measurement of the circulating levels of EDAc-MPs, EDAp-MPs, PDAc-MPs, PDAp-MPs predicted prognostic outcomes in advanced NSCLC patients.

NSCLC patient enrollment and therapeutic strategies
Patient enrollment, data collection, classification of advanced stage NSCLC and therapeutic strategies were according to our previous studies [18][19][20]. We assessed images and pathological findings of all patients who received evaluation or treatment for LC at Kaohsiung Chang Gung Memorial Hospital. We determined the eligibility of patients for interventions including surgery, adjunctive or palliative chemotherapy, irradiation therapy and/or target therapy based on the AJCC cancer staging criteria, 7 th edition [39]. Based on the radiological findings, we categorized LC patients into stages I, II, IIIA, IIIB, and IV according to AJCC cancer staging criteria, 7 th edition [42]. Patients with stage IIIB or stage IV NSCLC were categorized as advanced stage NSCLC. All the patients were enrolled for further evaluation, blood sampling and treatment in the outpatient department or upon hospital admission.
Detailed in-hospital and follow-up data including age, gender, chest x-ray findings, computed tomography, fibro-bronchoscopic findings, bone scans or ultrasound studies, other image findings, histological, pathological and laboratory findings were collected prospectively and entered into a computer database for analyses.
Informed consent was obtained from all patients and control subjects enrolled in the study. The study protocol was approved by the Institutional Review Committee on Human Research at Kaohsiung Chang Gung Memorial Hospital (IRB number: 100-1024B). The clinical investigations were conducted according to the principles outlined in the Declaration of Helsinki.
To circumvent adverse influences on measurement of circulating level of MPs, patients with one or more of the following criteria were excluded based on our previous studies [18][19][20]: (1) recent surgery or trauma during the preceding 2 months;(2) refusal to participate in the study; (3) other co-existent or history of malignances; (4) severe organ disease other than LC like chronic kidney disease (CKD > stage III), liver cirrhosis, hematologic disorders, congestive heart failure; (5) current use of anti-platelet agents;(6) history of febrile disorders; (7) acute or chronic inflammatory disease other than LC during the study period; or (8) (Table 3). These 136 patients were further divided into disease controlled (DC) group (n=94) and disease progression (DP) group (n=42).

Blood sample collection and processing
To determine circulating levels of MPs in advanced stage NSCLC patients, blood samples were collected at 9:00 am prior to and at the end of the first and third month after therapeutic interventions according to previously published protocol [18][19][20]. Additionally, blood samples were also collected at 9:00 am once from control subjects.
For flow cytometry, peripheral blood was collected in acid citrate dextrose (ACD) vacutainer tubes. Plateletrich plasma was prepared by centrifuging 1.5ml peripheral blood at 2500 g at 4°C for 15 min without acceleration. Then, 250μl plasma samples were thawed and centrifuged at 19,800 g for 10 min at 4°C, and then collected for analyzing MPs smaller than 1.0μm.
Size calibration was conducted with 1.0μm beads (Invitrogen, Carlsbad, CA). The MP pellet was re-suspended with 150μl of AnnexinV binding buffer (BD Biosciences). All buffers were sterile-filtered with a 0.2μm filter. Then, 100μl MPs were incubated in a TruCOUNT tube (BD Biosciences) with the following fluorescent monoclonal antibodies: (1) phycoerythrin (PE)-conjugated anti-CD31 (BD Biosciences); (2) fluorescein isothiocyanate-conjugated anti-AnnexinV (BD Biosciences) and; (3) phycoerythrin-Cy5 (PE-Cy5)conjugated anti-CD42b (BD Biosciences). The samples were incubated in the dark for 15 min at room temperature followed by addition of 400μl AnnexinV binding buffer and then analyzed in a FC500 flow cytometer (Beckman Coulter). The absolute count of MPs was measured by setting up the FACS machine with TruCount beads at 10,000 events. Additionally, white blood cell (WBC) counts, biochemistry and electrolyte levels were analyzed by standard laboratory methods in our hospital.

Disease classification
Change in tumor burden was assessed to determine tumor response to adjunctive therapy [18][19][20]. The chest computed tomography (CT) scans were routinely performed at baseline and at 12 week intervals after adjunctive therapy to determine the status of the disease. The tumor measurement was based on the current guidelines of Response Evaluation Criteria in Solid Tumors (RECIST) including complete response, partial response, stable disease and progressive disease [39]. Accordingly, we categorized the disease status as diseasecontrolled (DC) or disease-progressed (DP). The DC status was determined at the 3rd month after the treatment and defined as disease with regression with complete response, partial response or stable disease. On the other hand, the DP was defined as disease unresponsive to therapy with a growing tumor or metastasis after complete course of treatment. To elucidate if the changes in MPs over timecourse of treatment predicted outcomes, the relative levels (initial vs. third month) of the four types of microparticles were analyzed. The differences in the four types of microparticles were designated as ΔEDAp-MPs, ΔEDAc-MPs, ΔPDAp-MPs and ΔPDAc-MPs, respectively.

Statistical analysis
Data were expressed as means ± standard deviation (SD). Continuous variables were analyzed by independent t tests and categorical variables were analyzed by the chi-square test. To determine outcomes, we compared levels of MPs three months after treatment with the initial levels. All variables were considered as risk factors with a P < 0.10 in univariate analysis and were further analyzed by the multivariate logistic regression analysis to identify the independent factors that predict progressive disease. Receiver operating characteristic (ROC) curves were plotted and the area under the curve and CEA levels were compared for the four types of MPs. The cutoff value of MPs for predicting progressive disease in NSCLC patients was according to ROC curves. Results were presented as absolute numbers (percentage) or mean ± SD as well as medians, interquartile ranges and ranges for the various MP results. A two-tailed P value of less than 0.05 was considered statistically significant. Statistical analysis was performed using SPSS statistical software for Windows version 13 (SPSS for Windows, version 13; SPSS Inc., IL).