A pilot trial to examine the association between circulating endothelial cell levels and vascular injury in patients with diabetes and chronic kidney disease

Study design and setting

This was designed as a cross-sectional study of CEC levels and other vascular injury markers in patients with T2DM and CKD. Recruitment occurred at the outpatient nephrology offices of a large northeastern hospital. This office has an annual caseload of approximately 1500 patients with T2DM and CKD, of whom 25% are minorities (black or hispanic). The study was approved by the local institutional review board (local board reference number 13002, IRBNet ID – 403557), and all patients gave their written informed consent prior to study procedures.

Study population

Patients aged 18–89 years with T2DM for ≥ 8 years, and stage 3 CKD with an estimated GFR between 30 and 60 mL/min/1.73m² were considered for study inclusion. This inclusion criterion selected for a population at high risk for vascular injury that had not yet progressed to advanced CKD. Patients with prior evidence of renal artery stenosis, biopsy proven non-diabetic etiology of kidney disease, recurrent urinary tract infections, microhematuria, active malignancy, recent hospitalization and nephrotic range proteinuria were excluded. The exclusion criterion selected for patients with the highest likelihood of having diabetes related renal vascular injury, minimizing the possibility of elevations in CEC levels caused by alternative etiologies. Patients were selected to ensure approximately equal numbers of males and females, and normoalbuminuric and albuminuric subgroups.

Twelve hundred electronic charts with the ICD 9 coding diagnoses of T2DM and stage 3 CKD were selected for review of inclusion/exclusion criteria. Of these charts, 800 patients met all inclusion/exclusion criteria and 103 consecutive patients were asked to participate in the study when they presented for a follow-up appointment at our outpatients offices. Of those patients who were approached, 40 (38%) signed consent forms and participated in a single study session apart from their regularly scheduled follow-up visit.

Study procedures

All study procedures were performed during a single, morning 1-hour study session. For all patients, a complete history and physical including seated blood pressure was performed and baseline characteristics were recorded. For blood pressure readings, the lower of two seated measurements taken after 5 minutes of sitting was recorded. Comorbidities were defined by patient report or documentation in the patient’s problem list. Cardiovascular disease was defined as a prior history of coronary artery disease (CAD), congestive heart failure (CHF), peripheral vascular disease (PVD) or stroke.

Blood and urine samples were collected from all patients at the time of recruitment. This included serum creatinine, uric acid, von Willebrand Factor antigen levels, high sensitivity C-reactive protein, morning urine microalbumin to creatinine ratios (UACR) and CEC levels. For low density lipoprotein (LDL) cholesterol, hemoglobin A1c (HbA1c), serum phosphorus and blood hematocrit, the most recent value from outpatient blood work taken within 6 months of the study visit was recorded. Patients with UACR < 30mg/g were classified as normoalbuminuric, while patients with a UACR ≥ 30mg/g were classified as albuminuric.

Circulating endothelial cell measurement

Levels of total CECs were measured by the Clinical Lab Services Group at Janssen Diagnostics, LLC (previously Veridex Corporation) using the CELLSEARCH® Circulating Endothelial Cell Kit (Janssen Diagnostics, LLC). This system has provided valid and reproducible CEC levels in normal populations and in patients with chronic illness and malignancy, and the details have previously been described in detail^19,20. Briefly, peripheral whole blood was obtained with a 21-G butterfly needle and the first 3mL was discarded to minimize falsely elevated CEC levels related to venipuncture¹⁹. Whole blood was then collected into a tube containing anticoagulant and preservative, maintained at 15–30°C and shipped overnight to the central laboratory where it was processed within 72 hours. Prior studies have shown no significant difference in CEC levels at 0, 24, 48 and 72 hours in normal patients¹⁹. The CEC levels in a normal healthy population from this study are quoted from previous studies performed using the CELLSEARCH® Circulating Endothelial Cell Kit (Janssen Diagnostics, LLC)¹⁹.

CECs were isolated and enumerated by Janssen Diagnostics, LLC using the CELLSEARCH® Circulating Endothelial Cell Kit (Janssen Diagnostics, LLC) in combination with the CELLTRACKS® AUTOPREP® System (Janssen Diagnostics, LLC), and the CELLTRACKS ANALYZER II® (Janssen Diagnostics, LLC). These systems use immunomagnetic bead technology and a semi-automated fluorescent microscope. Briefly, 4mL of blood were mixed with 10mL of dilution buffer containing 0.1% sodium azide and the mixture was centrifuged at 800g for 10min, placed in the sample preparation system, i.e. the CELLTRACKS® AUTOPREP® System (Janssen Diagnostics, LLC), and processed within 1 hour of preparation. The sample preparation system then aspirated the plasma, leaving the buffy coat layer and red cells to be further processed. An anti-CD146 ferrofluid from a 3.0mL suspension of 0.012% magnetic nanoparticles conjugated to a mouse monoclonal antibody was added (CELLSEARCH® Circulating Endothelial Cell Kit; Janssen Diagnostics, LLC). The cells were then incubated inside the sample preparation system at room temperature for 12 to 13 minutes, at which time they underwent magnetic separation. The unbound cells and remaining plasma were then aspirated by the sample preparation system, and staining reagents with a permeablization buffer, to fluorescently label immunomagnetically bound cells, were added. These staining reagents included: 1) 4,6-diamidino-2-phenylindole (DAPI), a nuclear stain, 2) the mouse monoclonal CD105-PE which is specific for the protein endoglin and is expressed by endothelial cells; activated monocytes, stromal cells and pre-B cells, and 3) the mouse monoclonal pan-leukocyte antibody CD45 conjugated to allophycocyanin, in conjuction with a permeabilization buffer (CELLSEARCH® Circulating Endothelial Cell Kit; Janssen Diagnostics, LLC). Cells were then re-incubated and magnetic separation was repeated. Finally, the cells were re-suspended in approximately 300 µL of buffer and transferred to a chamber with two magnets that magnetically oriented cells into a monolayer for analysis. Analysis was performed within 24 hours using the CELLTRACKS ANALYZER II (Janssen Diagnostics, LLC), a four-color semi-automated fluorescent microscope and a gray scale charged coupled device camera, capturing 175 frames per channel. The captured frames were analyzed by image analysis software. CECs candidates selected by the software were then manually identified by the operator as DAPI+, CD105+ and CD45- cells. A single CEC count was performed for each blood specimen.

Statistical analysis

The aim of this pilot trial was to study the feasibility of recruitment, and assessment procedures. The sample size of 40 was based on the feasibility of single center recruitment and the study budget. All normally distributed continuous variables are reported as mean±standard deviation, and all non-normally distributed variables as median (1^st quartile – 3^rd quartile). Categorical variables are reported as proportions. Kolmogorov-Smirnov test was used to evaluate the normality of all continuous variables. A dichotomous grouping variable was created using urine microalbumin variable (‘albuminuric’ if urine microalbumin >30 mg/g, ‘normo-albuminuric’ if <=30 mg/g). Two independent samples t-test was used to compare normally distributed variables between albuminuric and normo-albuminuric groups and Wilcoxon rank-sum test was used for non-normally distributed variables. Fisher’s exact test was used to compare all categorical variables. Correlations analyses between levels of CECs and other vascular injury markers were performed using Pearson’s correlation for normally distributed continuous variables and Spearman’s for non-normally distributed continuous variables. There were three missing data points in this study (von Willebrand factor antigen × 1, low density lipoprotein X 2). This missing data was ignored from analysis. All analyses were performed using SAS 9.4. Results were considered significant if p<0.05.

Baseline characteristics

The characteristics of our study population are listed in Table 1. The average age of our group was 69.7 years, 52% were female, 72% were white and 20% were black (Table 1). A history of cardiovascular disease was present in 43% of participants, with a history of CAD in 35%, CHF in 8%, PVD in 10%, and stroke in 0%. Mean levels of all vascular injury markers were significantly higher than the reference range for our laboratory. Characteristics of albuminuric and normoalbuminuric subgroups are listed in Table 1. Only diabetes duration was significantly different between subgroups.

Circulating endothelial cell levels in normoalbuminuric and albuminuric groups

The mean CEC level in our study patients was 38.7±38.1 cells, range 7–196 (Figure 1). These levels are significantly higher than levels tested in a historical normal population. In that study of 249 healthy subjects, done by Immunicon Corporation using the CELLSEARCH System®, the mean CEC level was 21±18 cells, range 0–97¹⁹. In our albuminuric subgroup, the average CEC level was 34.8±34.3 and in the normoalbuminuric subgroup the average CEC level was 42.9±42.5 (Figure 1). There was no significant difference in CEC levels between the albuminuric and normoalbuminuric groups (p=0.297).

Figure 1. CEC levels of T2DM patients with CKD by albuminuria status.

Box plots shows 25th, 50th (median) and 75th percentiles (horizontal bars). The lower fence is 1.5 × inter-quartile range (IQR) below the 25th percentile. The upper fence is 1.5 × IQR above the 75th percentile. The circle and plus signs inside the boxes are means. The circles outside the fences are outliers. Abbreviations: CEC, circulating endothelial cells; T2DM, type 2 diabetes mellitus; CKD, chronic kidney disease.

Correlation between circulating endothelial cell levels and other vascular injury markers

Table 2 lists the correlation between CEC levels and other vascular injury markers tested in our study. We found a significant correlation between CEC levels and uric acid levels (r=0.33, p=0.039). No other significant correlations were found.