Background

Sleep apnea syndrome (SAS) is a sleep disturbance, which is frequently comorbid in dialysis patients [1, 2]. SAS induces hypoxia during apnea, and therefore, systemic and cerebral oxygenation would be low in these patients. However, few reports have examined regarding the changes in cerebral oxygenation during sleep in patients with SAS. Near-infrared spectroscopy (NIRS) has recently been used to measure regional saturation of oxygen (rSO2), as a marker of tissue oxygenation. In particular, cerebral rSO2 was measured in patients in various clinical settings, including hemodialysis (HD) patients [3,4,5,6,7,8]. It would be considered important to maintain cerebral rSO2 levels because that low cerebral rSO2 has been associated with cognitive impairment in patients with chronic kidney disease (CKD) [5, 8] and poor neurologic outcomes in post-cardiac arrest patients [7]. We aimed to examine the associations between changes in cerebral rSO2 and those in systemic oxygenation using the measurement of peripheral arterial oxygen saturation (SpO2%) in a patient undergoing HD with SAS.

Case presentation

A 74-year-old woman underwent HD therapy for 1 month, because of advanced CKD caused by hypertension and obesity. In addition, she was diagnosed with obstructive sleep apnea (OSA) with polysomnography about 4 years prior. Her apnea-hypopnea index (AHI) was 77.5 per hour and the arousal index was 76.1 per hour. Therefore, continuous positive airway pressure (CPAP) was started, and her AHI was improved to 3 to 6 per hour. However, she discontinued CPAP therapy at HD initiation. As systemic oxygenation would deteriorate without CPAP, we assessed her overnight SpO2 using a pulse oximeter (PULSOX-Me300; TEIJIN PHARMA, Tokyo, Japan). Additionally, for the evaluation of cerebral oxygenation, we monitored cerebral rSO2 using a regional saturation monitor (INVOS 5100c; Covidien Japan, Tokyo, Japan). Her vital sign and laboratory findings were as follows: blood pressure (BP) and pulse rate (PR) before evaluation; 127/71 mmHg, 76 beats/minute, BP and PR after evaluation; 121/59 mmHg, 80 beats/minute, hemoglobin (Hb): 9.5 g/dL, serum albumin (Alb); 2.7 g/dL, blood urea nitrogen (BUN): 48.2 mg/dL, serum creatinine (Cr): 4.13 mg/dL. Her physiological function examination findings were as follows: ejection fraction (EF): 60%, left ventricular systolic function was normal, mitral valve regurgitation was mild by echocardiography, and left and right carotid max intima-media thickness (IMT) were 1.1 and 1.1 mm, respectively by carotid ultrasonography. Cerebral rSO2 level was evaluated as the rSO2 ratio, which was defined as the ratio of rSO2 at t (min) during sleep to the initial rSO2 before bedtime [rSO2 at t (min) during sleeping/initial rSO2 before bedtime]. We confirmed the deterioration in cerebral rSO2 according to the SpO2 decrease in this HD patient with SAS. However, we could not directly comment on the influence of CPAP therapy on the improvement in systemic oxygenation because she refused to receive CPAP therapy after HD initiation. Therefore, as a comparison, changes in SpO2 and cerebral rSO2 were evaluated in a 69-year-old male HD patient without SAS. His vital sign and laboratory findings were as follows: BP and PR before evaluation 135/92 mmHg, 68 beats/minute, BP and PR after evaluation 148/100 mmHg, 61 beats/minute, Hb 11.1 g/dL, Alb 3.1 g/dL, BUN 55.7 mg/dL, and Cr 7.86 mg/dL. Results of his physiological function examination were as follows: EF: 71%, left ventricular systolic function was normal contraction, no valvular disease was noted by echocardiography, and left and right carotid IMT was 3.2 and 2.4 mm, respectively by carotid ultrasonography. Both patients provided informed consent before evaluation, and these examinations in each patient were performed on the day without HD therapy. The SpO2 gradually decreased in patients with SAS from initiation to 3 h after sleep and then fell below 90%. According to the SpO2 decrease during sleep, her cerebral rSO2 simultaneously decreased; furthermore, the improvement in cerebral rSO2 was apparently delayed even after the improvement in her SpO2 (Fig. 1a). Her 3% Oxygen Desaturation Index was 19.0 times/hour by pulse oximetry. However, as shown in Fig. 1b, both his SpO2 and cerebral rSO2 ratio were maintained throughout sleep in the HD patient without SAS.

Fig. 1
figure 1

Time course of SpO2 and cerebral rSO2 in a patient with obstructive SAS (a) and a patient without SAS (b). SpO2 peripheral arterial oxygen saturation, rSO2 regional saturation of oxygen, and SAS sleep apnea syndrome

Discussion

We experienced a HD patient with SAS who showed a decrease in cerebral oxygenation in addition to a decrease in systemic oxygenation during sleep. In general, a decrease in systemic oxygenation would induce the deterioration of tissue oxygenation, including the brain. However, thus far, few reports have directly examined changes in cerebral oxygenation during sleep. Therefore, this case is a first report about nocturnal real-time monitoring of cerebral rSO2 in a HD patient with SAS.

Sleep disorders, including SAS, are known as long-term complications in dialysis patients [1, 2]. Nocturnal hypoxia was associated with a poor prognosis for cardiovascular events [9], and patients with ischemic stroke have a high risk of obstructive sleep apnea, with lowest overnight SpO2 values < 80% in more than half of observed patients [10]. Furthermore, in the clinical setting of HD therapy, cognitive impairment is common in patients undergoing HD and the pathogenesis of cognitive impairment would be associated with hemodynamic factors including changes in BP or anemia [11]. In particular, both intradialytic hypotension and anemia reportedly induced the deterioration of cerebral oxygenation [12, 13]. In addition, the decrease in cerebral oxygen supply was associated with neuronal death, and hypoxia promoted the formation of amyloid ß peptide in experimental studies [14, 15]. Therefore, systemic oxygenation should be maintained to prevent worsening of cerebral oxygenation, which would be associated with the deterioration of cognitive function, in patients with SAS.

The measurement of cerebral rSO2 by NIRS has recently been performed in various clinical settings, including cardiovascular surgery, pediatrics, and dialysis [3,4,5,6,7,8, 16,17,18]. We therefore, reviewed the literature for reports of cerebral rSO2 values in dialysis patients (Table 1). Cerebral rSO2 values in HD patients were lower than healthy control or non-dialysis patients [3, 8, 19,20,21] and were lower than peritoneal dialysis patients [22]. Therefore, the presence of renal dysfunction or dialysis therapeutic modality might influence the status of cerebral oxygenation. Furthermore, longer HD therapy itself might cause hypoxia of the brain because HD duration was negatively associated with cerebral rSO2 [3, 19]. Regarding the association between cerebral rSO2 and clinical parameters in maintenance HD patients, various factors, i.e., serum albumin, pH, Hb, and blood pressure, were related to cerebral rSO2 [3, 13, 23]. Based on these previous reports, it would be important to maintain the nutritional status, and control Hb level and blood pressure in the clinical setting of HD therapy. On the other hand, it was also reported that there are little changes in cerebral rSO2 during HD in spite of Hb increase induced by ultrafiltration in patients undergoing HD without intradialytic hypotension [6, 13, 20, 24]. However, thus far, there are few reports regarding the association between changes in systemic oxygenation and those in cerebral oxygenation in HD patients with or without SAS. Recently, evaluation of cerebral rSO2 was reported to be useful for detection of SAS in addition to that using a pulse oximeter in children [18]. This report described that cerebral NIRS monitoring could detect SAS even in patients in whom SAS was ruled out with polysomnography, and AHI determined using cerebral rSO2 had higher sensitivity for the detection of SAS compared with that using SpO2 [18]. In HD patients with SAS, we noticed that a decrease in cerebral rSO2 was observed more frequently than a decrease in SpO2, and this result might be consistent with a previous report [18]. On the other hand, the HD patient without SAS in this study showed little changes in SpO2 and cerebral rSO2 during sleep. Therefore, in addition to the usefulness of polysomnography evaluation as a standard method for diagnosing SAS, cerebral rSO2 measurement using NIRS might play an important role in the diagnosis of SAS in HD patients.

Table 1 Studies regarding cerebral oxygenation by near-infrared spectroscopy monitoring in dialysis patients

Furthermore, the improvement in cerebral rSO2 was apparently delayed even after the improvement in SpO2 level (Fig. 1a). Although it is difficult to accurately state the mechanism of this phenomenon, we suspected two reasons regarding the difference between systemic and cerebral oxygenation improvement. First, NIRS monitoring was reported to have higher sensitivity for detecting SAS [18]. Therefore, the delay in cerebral oxygenation improvement might have the possibility to reflect the deep tissue oxygenation status, which was different from superficial arterial oxygenation status as shown in SpO2 monitoring. Second, NIRS monitoring would mainly help observe venous oxygenation status (70–80%), against capillary (5%), or arterial blood (20–25%) [25]. On the other hand, SpO2 monitoring could observe in the arterial blood oxygenation status. Brain cell would need the uptake of steady amount of oxygen into the cell according to the changes in oxygen supply in various conditions. In situations of decreased oxygen supply to brain such as severe anemia or hypoxic condition in SAS, cerebral rSO2 would be low, which reflected the decrease of oxygen supply and the increase of oxygen extraction into the brain cell [13]. When hypoxic status in SAS transiently disappears during sleep, arterial oxygenation will be immediately improved, which lead to the SpO2 increase. However, changes in oxygen extraction into brain cell might not be immediately normalized in response to the sudden increase of oxygen supply. As a result, there will be a lag time between the improvement in cerebral oxygenation and that of systemic oxygenation during sleep in this SAS patient. However, the mechanism of this phenomenon is yet not to be accurately clarified, and therefore, further studies are needed.

The use of CPAP improves systemic hypoxia in patients with SAS, and patients with OSA may have better outcomes under CPAP therapy through a decrease in the rate of recurrent myocardial infarction [26]. According to her prior medical records, the patient’s AHI became relatively low under CPAP therapy compared to that without CPAP. Therefore, CPAP therapy would be effective in the prevention of systemic hypoxia during non-HD periods. Based on this result, cerebral oxygenation could be improved by CPAP therapy in patients with SAS in addition to the improvement of systemic oxygenation. However, further studies are needed to confirm the association between improvement in cerebral oxygenation and CPAP therapy in patients with SAS because we were unable to observe these associations in this study.

In conclusion, we observed cerebral oxygenation deterioration during sleep in addition to a decrease in systemic oxygenation in a patient undergoing HD with SAS. Furthermore, cerebral NIRS monitoring during sleep might be a useful method for the detection of SAS.