This study demonstrated that when the anesthesia was induced slowly and maintained without excessive decrease of MAP, SpO2, and BIS values and with adequate etCO2 adjustment, the nTHI (which reflects the CBF) decreased by 40% in elderly patients with severe dementia after the start of treatment, indicating significantly lower nTHI values than in elderly patients with no cognitive decline (−25%). No change was observed in the TOI in either group.
The NIRS oxygenation monitoring system utilizes the high tissue permeability of near-infrared light. Near-infrared light that enters body tissues is absorbed by oxygenated and deoxygenated hemoglobin molecules, which have different absorption spectra for near-infrared light. The spectral difference is used to determine the tissue oxygen saturation [36]. Factors that affect measured values include anemia, effect of extracranial blood flow, and effect of body position [36]. In this study, none of the patients had anemia with Hb < 10 g/dL, the probe was applied to an area of the forehead where there were no skin veins, and the measurements were taken after the body was in a horizontal position and rested for 5 min [36]. Therefore, it was considered that these effects could be eliminated.
The CBF is physiologically maintained constant in the range of MAP 60–150 mmHg under consciousness (autoregulation of CBF) [38]. The CBF increases 1 to 2 mL/100 g/min for each 1-mmHg increase in PaCO2 around normal PaCO2 values (PaCO2 range, 25–70 mmHg) [39]. Furthermore, CBF increases rapidly when PaO2 < 60 mmHg. Meanwhile, 60% of the brain's energy expenditure contributes to electrophysiological functions, and the cerebral metabolic rate (CMR) is related to CBF [39]. MAP and CBF (middle cerebral artery flow velocity) have been reported to be correlated during anesthesia in a study of patients with an average age of 51 years [40], and that in the anesthesia using propofol (1–2 mg/kg), remifentanil (anesthesia induction with 1 µg/kg/min and maintenance using 0.1 µg/kg/min), and sevoflurane (< 1 minimum alveolar concentration [MAC]), the MAP and degree of expansion of the internal carotid and middle cerebral arteries are correlated [41]. These reports suggest that MAP and CBF are correlated even under anesthesia. In this study, the lowest values set of vital sign during anesthesia (avoiding MAP < 50 mmHg and BIS < 20) were determined as a range that does not affect CBF based on these physiological data. In the present study, the lowest value of MAP was 61 mmHg (median value), which is considered to be within the category of autoregulation of CBF. PaCO2 of 35–40 mmHg (median value) and SpO2 of 96–100% (equivalent to PaO2 > 80 mmHg; median value) are considered to have small physiological effect on CBF. The lowest value of BIS is 43 (median value), which are appropriate for general anesthesia without suppressing the electrophysiological functions excessively.
When a volatile anesthetic is administered, CBF is determined by the balance between CBF reduction due to CMR suppression and CBF increase due to direct cerebral vasodilation. When the volatile anesthetic is 0.5 MAC, CBF decreases because CMR suppression becomes dominant. However, at 1 MAC, both are antagonistic and CBF does not change, but when it exceeds 1 MAC, vasodilation becomes dominant and CBF increases significantly [39]. In this study, despite the difference in sevoflurane concentration (median value) during anesthesia induction (3.0% in the dementia-positive group and 5.0% in the dementia-negative group), both groups had > 1 MAC, and MAP > 60 mmHg was maintained. Although CBF should be physiologically in a state of increase, both groups showed a decrease in CBF. Furthermore, during anesthesia maintenance, the sevoflurane concentration (median value) in both groups was 0.5–1.5%, < 1 MAC, and MAP > 60 mmHg; thus, physiologically, the CBF should remain unchanged. However, again, both groups showed a decrease in CBF.
Regarding intravenous anesthetics and narcotics, during anesthesia induction, fentanyl 75 µg (median) and remifentanil 0.1 µg/kg/min were used in the dementia-positive group, and fentanyl 75 µg and remifentanil 0.25 µg/kg/min in the dementia-negative group. During maintenance of anesthesia, the dementia-positive group received 25 µg of fentanyl and 0.05–0.1 µg/kg/min of remifentanil, and the dementia-negative group received 50 µg of fentanyl and 0.1–0.15 µg/kg/min of remifentanil, with no significant difference. Propofol, which was used in the pilot study, decreased CBF by 53–79%, potentially affecting the results, and it was therefore not used in the present study [39, 42]. Some studies reported no change in CBF and cerebral metabolic rate of oxygen (CMRO2) when fentanyl was used at an average dose of 16 µg/kg, and other studies reported that CBF increased in the frontal, temporal, and cerebellum areas when used at 1.5 µg/kg, but decreased in other areas. Meanwhile, remifentanil at 0.05–0.15 µg/kg/min causes a mild increase in CBF in the frontal, inferior parietal lobule, and motor cortex region, and a decrease in the cerebellum, superior temporal gyrus, and periaqueductal gray matter [39, 43]. As the CBF was measured in this study using a probe on the forehead, both fentanyl and remifentanil at the doses used in this study should physiologically indicate a slight increase in CBF on the forehead. However, both groups showed a decrease of CBF.
In a report on anesthesia and CBF in patients without dementia, Soejima et al. reported that with anesthesia using remifentanil (0.1–0.3 µg/kg/min) and sevoflurane (1–2%), MAP decreased to approximately 60 mmHg (−25%) and CBF was −20% from the baseline, consistent with the present study [44]. The following factors concerning reduction of CBF are considered: MAP has decreased to the lower limit of autoregulation of CBF; CBF may have decreased due to propofol [44]. In other studies in patients with an average age of 47 years, anesthesia using propofol and remifentanil decreased both MAP (−17–34%) and CBF (blood velocity of the middle cerebral artery: Vm, −28–39%) [45], and another study showed that CBF decreased to 81% at 0.75 MAC of sevoflurane, 74% at 1 MAC, and 79% at 1.25 MAC compared with baseline (before sevoflurane administration) [46]. Based on these results, CBF decreases by approximately 20–25% due to general anesthesia. Even in elderly people older than 65 years, the upper and lower limits of MAP at which CBF becomes constant do not change at a sevoflurane concentration of 1 MAC [47], and in another study of patients with an average age of 37 years, CBF autoregulation and CO2 reactivity are maintained at a sevoflurane concentration of 1 MAC [48].
Even in the patients without cognitive decline in this study, the CBF decreased by −25% due to anesthesia, which is similar to a previous report [44]. In this study, there was less significant difference in the dose of anesthetics or opioids between the two groups. The CBF should physiologically increase with sevoflurane of 1 MAC or more at the time of anesthesia induction, or CBF should stay constant during anesthesia maintenance with 0.5 MAC of sevoflurane, fentanyl and remifentanil. However, it is unlikely that CBF decreased. Furthermore, because the type and amount of anesthetics used were adjusted in this study, both elderly patients with dementia and those without cognitive decline showed similar trends in the MAP, BIS, SpO2 and etCO2 values. Therefore, when MAP decreased to the lower limit of autoregulation of CBF in the patients aged approximately 70 years, CBF could decrease in both groups. However, the degree of decline is greater in patients with severe dementia; elderly patients with severe dementia might have had a weakened CBF autoregulation mechanism or already decreased brain tissue metabolism.
Herein, no changes in the TOI were found in either the dementia-positive or dementia-negative group. Although the TOI does not change as rapidly as arterial oxygen partial pressure and SpO2, a study reported an increase in the TOI with oxygen administration [37]. In this study, TOI may not have changed in the present study, as oxygen was administered at 100% during anesthesia induction and 33% during anesthesia maintenance.
The highest MAP value after induction of anesthesia was significantly higher in the dementia-positive group (MAP 3: median, 98 mmHg) than in the dementia-negative group (MAP 3: median, 84.5 mmHg). The highest BIS value also showed significantly higher values in the dementia-positive group (BIS 2, 3, and 4: median values, 75, 67, and 70, respectively; non-dementia group: median values, 58, 57, and 55, respectively). These values in the dementia-positive group may have increased due to stimulation such as anesthesia induction and tracheal intubation because the depth of anesthesia was somewhat shallow in the dementia-positive group. Because anesthetics and narcotics can decrease the blood pressure and BIS values excessively in elderly patients, we limited the amount used [39].
As the dementia grades (FAST and CDR) of the patients with dementia did not deteriorate after anesthesia, we concluded that the nTHI values decreased, but there was no decline in cognitive function among patients with severe dementia at FAST stages 6e to 7c and/or CDR indicating severe dementia. However, each category of dementia grade (FAST and CDR) has a certain range of symptoms. A subtle cognitive decline might result in the patient continuing to be classified in the same grade, due to the weak detection capability of FAST and CDR [34].
The limitation of this research is that the probes for measuring the brainwaves (BIS) and the CBF (NIRO®-200NX) had to be placed on the forehead, and although the BIS and CBF at the forehead could be measured, the conditions at other locations in the brain were not evaluated. Some reports showed differences in blood flow changes in different parts of the brain depending on the anesthetics used; thus, more detailed measurement at the multiple sites of the brain may be necessary [39]. In the present study, we also tried to adjust the types and volumes of anesthetics and opioids in both groups in order to maintain the MAP, BIS, SpO2, and etCO2 values within a preferable range. However, complete matching of the drugs used and their volume was not possible. Although the anesthesia method used in this study is thought to have a small effect on CBF in both groups, it may be an issue that should be considered when discussing the results. Although this study satisfied the requirements of the statistical sample size calculation, the sample size was small, which is another limitation. Future large-scale studies should evaluate cerebral and cognitive function during and after anesthesia.
The present study showed that when anesthesia was slowly induced while controlling the MAP, BIS, SpO2 and etCO2 values within a preferable range, the nTHI value of elderly patients with severe dementia was 0.60 (decreased by 40%); however, an nTHI value of 0.75 (decreased by 25%) was observed in elderly patients with no cognitive decline. These findings indicate that the nTHI value decreases significantly in elderly patients with severe dementia. No change was observed in the TOI.