The peripheral neutrophils in subjects with COPD-OSA overlap syndrome and severe comorbidities : A feasible inflammatory biomarker ?

Material and methods. A cross-sectional study of patients with COPD and severe comorbidities, as defined by a Care Assessment Need score over 95, were seen in the Pulmonary Tele-Health Clinic at the Salem Veteran Affairs Medical Center, USA, over a 1-year period. Demographic and polysomnographic data, FEV1 and the Charlson Comorbidity Index (CCI) were extracted from the Electronic Medical Records. Obstructive sleep apnea was defined according to the American Academy of Sleep Medicine (AASM) guidelines. Serum inflammatory markers (PPN, CRP, fibrinogen and procalcitonin) were obtained after the Tele-Health appointment.


Introduction
Chronic obstructive pulmonary disease (COPD), a disease characterized by irreversible airflow limitation, and obstructive sleep apnea disease (OSA), a disease characterized by intermittent collapse of the upper airway, are each responsible for major morbidity, mortality and economic burden. 1 The term "overlap syndrome" (OS) was introduced to describe the association of both conditions in a single patient.Given the high prevalence of both COPD and OSA, it is expected that up to 29% of OSA subjects could be affected by OS. 2 Although the evidence suggests that the increased mortality in OS subjects is primarily cardiovascular, the exact mechanisms remain unclear. 1 It is likely, though, that cell-mediated inflammation plays an important role given the shared systemic inflammation pathways in COPD, OSA and atherosclerosis. 3][6] As a clinical biomarker, increased neutrophil count correlates with the severity of coronary artery stenosis and survival in subjects with stable coronary artery disease and demonstrates good specificity as a prognostic marker for acute coronary syndrome when the electrocardiogram was nonspecific. 7,8McNicholas suggested that COPD and OSA are systemic disorders that can cause or worsen CVD, one of the mechanisms related to the chronic intermittent hypoxia that stimulates the inflammatory cytokines in which neutrophils participate. 3Despite these findings, the abovedescribed findings were not demonstrated in OS and little is known about the role of neutrophils in the underlying mechanisms of increased mortality in these subjects.We therefore aimed to assess peripheral neutrophils in OS subjects compared to COPD-only subjects with a similar degree of comorbidity.In addition, as it is has been suggested that the current standard therapy for OSA, i.e., continuous positive airway pressure (CPAP), could alter the inflammatory markers in OSA, we evaluated the inflammatory response in both CPAP-compliant and noncompliant OS subjects as compared to COPD-only subjects.

Material and methods
The study is a cross-sectional retrospective evaluation of subjects with COPD and severe comorbidities selected from subjects followed for routine primary care at the community-based outpatient clinics (CBOC) satellite to the Salem Veterans Affairs Medical Center (VAMC), USA, over 1 year.The severity of comorbidity was assessed by the Care Assessment Need Score (CAN), which indicates how a given patient compares to other individuals in terms of the likelihood of hospitalization or death.Higher CAN scores reflect an increased number of healthcare providers, medications and mental health diagnoses. 9Subjects with a CAN score greater than 95 were eligible for the study and their charts were screened for a pulmonary function test (PFT) that showed COPD according to American Thoracic Society (ATS) guidelines.Three hundred and fifty-three subjects with a CAN score over 95 were identified in the CBOC; 150 of these subjects were diagnosed with COPD.Thirty-eight of the COPD subjects were evaluated in the Pulmonary Tele-Health Clinic, out of which 17 had a diagnosis of OSA prior to this study and were defined as cases of OS.Demographic, clinical and medication data (inhaled and oral corticosteroid) were extracted from the Electronic Medical Records (EMR).The Charlson Comorbidity Index (CCI) and the Framingham (FH) 10-year cardiovascular risk score were calculated based on EMR data. 10,11Dyspnea severity was assessed using the Medical Research Council Scale. 12The study was approved by the Salem VAMC Review Board as quality improvement data.
Among the COPD participants, a chart review was performed for a diagnosis of OSA.The inclusion of OSA subjects was done based on a diagnostic polysomnogram completed prior to our study (SensorMedics, Sunnyvale, USA).The stages of sleep were identified using 2-channel electroencephalogram (C4-A1, C3-A2), chin electromyogram, and left and right electrooculograms.Thoracoabdominal movements were monitored with thoracic and abdominal strain gauges.Airflow was monitored with an oronasal thermistor per the clinical protocol.Arterial oxyhemoglobin saturation was recorded with the use of a pulse oximeter.Electrocardiogram, snoring and body position were also recorded.Recordings were manually scored according to the standard criteria. 13An episode of obstructive apnea was defined as an absence of airflow for at least 10 s, in the presence of rib cage and abdominal excursions.Hypopnea was defined as a discernible reduction in airflow lasting 10 s or longer and associated with at least a 4% decrease in arterial oxyhemoglobin saturation, an electroencephalographic arousal, or both.The number of episodes of apnea and hypopnea per hour was referred to as the apnea hypopnea index (AHI).The mean and minimal oxygen saturation for the entire night was estimated, along with the duration of stages 1, 2, 3, and rapid eye movement (REM) of sleep; sleep fragmentation was assessed using the arousal index (AI).The sleep parameters were scored and OSA was defined according to the criteria established by the American Academy of Sleep Medicine Manual for Scoring Sleep and Associated Events. 13orning blood samples were collected for white blood cell count (WBC) with the percentage of peripheral neutrophils (PPN), C-reactive protein (CRP), fibrinogen, and procalcitonin, from subjects who attended the Tele-Health appointment.Assays were performed in a single run by the Salem VAMC laboratory.
The normality of continuous variables was evaluated by the Shapiro-Wilk test.Comparison of continuous data between the OS and COPD-only groups was performed using a t-test for data with normal distribution.Peripheral neutrophils, systolic blood pressure and serum procalcitonin had a skewed distribution and the difference between groups was tested by the Wilcoxon-Mann-Whitney test.Since BMI has a known association with OS and chronic inflammation, stratified analysis by obesity (BMI >30 kg/m 2 ) was performed to assess for potential confounding.A 2-sided p-value <0.05 was considered statistically significant.All analyses were performed using STATA v. 11.0 software (Stata Corporation, College Station, USA).

Results
Thirty-eight subjects with confirmed COPD were included in the study and 17 (44%) of these were classified as OS.Seven out of the 17 subjects with OS had an excellent CPAP compliance as defined by Medicare guidelines, i.e., the use of CPAP devices for 4 or more hours per night on 70% of nights. 14The subjects' baseline characteristics are detailed in Table 1.The 112 subjects who were eligible for the study (CAN score >95 and a diagnosis of COPD) but could not be reached by phone or who declined participation in the Tele-Health appointments were not demographically different from those included in the analysis.
All OS subjects had an AHI greater than 5, except 1 patient whose AHI was 4.7.This patient was diagnosed with OSA and therefore classified as OS as well, due to the evidence of snoring with arousals during REM sleep, desaturations (at least 4%) and symptoms suggestive of OSA, such as excessive daytime sleepiness.The AI, as well as mean and lowest oxygen saturations during the sleep study are described in Table 2.There was a significant difference (p = 0.002) in PPN values in OS patients (73 ±9.3) vs COPDonly patients (60 ±14.9), regardless of whether OS subjects were compliant with the CPAP therapy (p = 0.03) or not (p = 0.005).There was no difference in the PPN values between CPAP-compliant and noncompliant OS subjects (p = 0.91).The association of PPN and OS remained significant for obese subjects (p = 0.04) and non-obese subjects (p = 0.03).The serum inflammatory markers, WBC, CRP, fibrinogen, and procalcitonin were similar between the OS subjects and COPD-only subjects (Table 3), regardless of their compliance (p > 0.11) or noncompliance (p > 0.17) with the CPAP therapy.No differences in the severity of COPD, baseline comorbidities, smoking, or home oxygen therapy (HOT) use (at night, during the day or continuously) were found between OS subjects and COPD-only subjects (Table 1).Body mass index was higher in the OS group (Table 2).Body mass index, COPD severity, smoking, and HOT use (diurnal, nocturnal or continuous) were not associated with PPN (p-values >0.2).

Discussion
In a clinic-based sample of community-dwelling subjects with a high level of overall comorbidity, we found that PPN is higher in OS subjects compared to COPD-only subjects, regardless of their compliance with the CPAP therapy.The severity of COPD was similar between groups and, moreover, was not associated with PPN.These results suggest that the additive adverse effect of OSA is associated with increased PPN in OS and they support further research on neutrophil-related inflammation and cardiovascular disease (CVD) in these subjects.
The role of increased inflammation in sleep disordered breathing (SDB) has been receiving increased attention recently.6][17] Data on the inflammatory response in OS is currently limited to only a few studies to date.Nural et al. demonstrated that the serum CRP level was higher in the OS group than the COPD-only or OSAonly groups. 18Mansour et al. showed that nasal CPAP decreased levels of TNF-α, CRP and IL-6 in these subjects. 19While no direct link has yet been demonstrated between inflammation and CVD in OS subjects, Marin et al. reported a higher number of cardiovascular deaths in subjects with untreated OS compared to treated OS, and a higher number of deaths than in those with COPD only. 1 Although the concept of inflammation-related CVD in SDB is largely accepted,the evidence supporting it is difficult to find as the typical inflammatory markers, such as CRP and IL-6, are heavily confounded by obesity and smoking status. 20Therefore, other serum markers, such as neutrophils, are worth exploring given their lack of a strong connection with BMI (less than 10% of the variance of change). 21Moreover, the neutrophil count is inexpensive, reliable, easy to interpret, and ordered routinely in inpatient and outpatient settings.Whether neutrophils are different in obesity or not is controversial, as both increased percentages in African-American children and unchanged numbers in men have been reported. 22,23In addition, Dixon et al. showed that even when neutrophils change with BMI in adults, this change was small and accounted for less than 10% of the variance of baseline counts. 21In our study, OS subjects had a significantly higher BMI than the COPD-only subjects, but BMI was not associated with PPN.
The detailed mechanism by which neutrophils contribute to inflammation-mediated CVD is yet unknown, but possibly related to an abnormal antiapoptotic response as a response to hypoxia, with increased neutrophil survival, migration towards endothelial injury and augmentation of the atherosclerotic nidus. 24Recent evidence suggests that hypoxia is associated with delayed apoptosis of neutrophils in COPD and decreases apoptosis in OSA subjects, a process that may induce vascular damage through the previously mentioned biochemical and hematological mechanisms. 24,25Overall, the apoptosis of neutrophils represents a fundamental and intricate process of intracellular death/survival signaling pathways whose overall balance aims at containing the inflammatory response. 26In OSA and COPD subjects, antiapoptotic myeloid cell leukemia 1 (Mcl-1) is upregulated and the proapoptotic Bax or Bak molecules are downregulated. 24,27This concept becomes especially important for OS subjects who have both COPD and OSA.9][30] These findings suggest congruently   that the study of neutrophils in OS is clinically significant given their participation in the pathophysiology of CVD and the minimal confounding effects of obesity.
To the best of our knowledge, our study is the first to compare PPN in OS subjects to COPD-only subjects, thus providing important clinical results that invite larger studies on neutrophil-related inflammation in these subjects.One benefit of the study is that both groups of subjects were followed in the same outpatient clinic and had a similar degree of comorbidity (CCI and FH score), smoking and corticosteroid use (inhaled and oral).The severity of COPD was not different between the groups and was not associated with PPN.
2][33] Drager et al. showed that the systemic inflammation in moderate to severe OSA was not influenced by short-term CPAP therapy. 32Although our study did not show a difference in PPN between CPAP-compliant and noncompliant OS subjects, it also had the same limitations as the other previous studies, i.e., a small sample size and variation in CPAP therapy duration and compliance.
Our study has a few other limitations.Firstly, we included subjects who were on HOT for chronic hypoxia or CPAP therapy for OSA.However, treatment with oxygen would bias the results against an association with PPN between the OS and COPD-only subjects.Secondly, since we used the CAN score as a screening tool, our study included a population that had other complex chronic illnesses and, as such, our results might not be applicable to all OS subjects.However, as the CAN score refers to the likelihood of hospitalization or death of a patient within a 1-year period, our results are applicable to the large population of high-risk OS subjects.In addition, our subjects were evenly distributed between the severity levels of COPD disease, with a majority of those having mild to moderately severe OSA, thus matching a large cohort of subjects with OS.Because we limited participants to those with COPD and a high CAN score, our study resulted in a small sample size.For this reason, these findings cannot be generalized to the broader community-based sleep subjects.Comparison to CRP was limited due to the lack of a high-sensitivity CRP (hs-CRP) assay.
In conclusion, our study of a small sample of subjects with COPD and severe comorbidities suggests that those with OS have higher PPN than those with COPD alone, regardless of their compliance status with the CPAP therapy.Future research in neutrophil-related inflammation and CVD in a larger and more diverse group of OS subjects is needed.If our results are confirmed in a larger study, they could be used to motivate OS subjects to improve their lifestyles (e.g., smoking cessation, dietary modification, exercise, and weight loss) or to comply with drug therapies aimed at reducing CVD.

Fig 1 .
Fig 1. PPN in Subjects with OS vs COPD-only.Group 0 = OS Subjects Non-Compliant with CPAP; Group 1= OS Subjects Compliant with CPAP; Group 2= COPD-only.PPN were higher in Subjects with OS than COPD regardless of the CPAP compliance status

Fig. 1 .
Fig. 1.Peripheral neutrophils (PPN) in subjects with OS vs COPD-only OS -overlap syndrome; CPAP -positive airway pressure therapy; COPD -chronic obstructive pulmonary disease; group 0 -OS subjects noncompliant with CPAP; group 1 -OS subjects compliant with CPAP; group 2 -COPD-only subjects; PPN was higher in subjects with OS than COPD regardless of CPAP compliance.

Table 2 .
Description of the severity of COPD and OSA disease

Table 3 .
Description of the inflammatory markers mean of v0 neutrophil