Study Setting, Design, Ethics, and Trial Registration

This nonfunded, single-center, parallel-group randomized controlled trial (RCT) was conducted at the National Institute of Mental Health and Neurosciences (NIMHANS), Bengaluru, after approval of the institutional ethics committee [NIMHANS/IEC (BS & NS DIV.) 13th meeting/2018 held on June 20, 2018] and written informed consent from the patients or their legal guardian. The trial was registered with the Clinical Trial Registry of India (CTRI/2018/07/014970, registered on July 19, 2018).

Study Population and Period

All consecutive consenting eligible patients with CCCM who were scheduled for decompressive surgery during the 15-month period from August 2018 to October 2019 were recruited in the study. Patients were included if they had clinical symptoms of myelopathy of ≥3 months with imaging evidence, belonged to American Society of Anesthesiologists (ASA) physical status 1 and 2, and aged between 18 and 70 years. We excluded patients with atlantoaxial dislocation and cervicomedullary junction pathology, tumors, and preexisting systemic disease such as diabetes mellitus, hypertension, ischemic heart disease, and chronic obstructive pulmonary disease.

Control of Bias and Study Interventions

Randomization was done using computer-generated random number table with 1:1 allocation ratio. Study interventions were performed in the preoperative holding area using identical 500-mL fluid bottle (infusion rate determined as per randomization) by an individual not directly involved in the study. Thus, patient, anesthesiologist, outcome assessor, and data analyst were blinded to the group allocation. The patients in group F received 20 mL/kg of Ringer's lactate over 30 minutes, while those in group C received Ringer's lactate at the maintenance rate of 2 mL/kg/h.

Data Collection

Demographic details such as age, gender, and body mass index were collected. The noninvasive cardiac output monitor (NICOM; Cheetah Medical, Tel Aviv, Israel) was used to obtain data regarding advanced hemodynamic parameters during the study period. The NICOM device works on the principle of bioreactance.12 Changes in the intrathoracic blood volume alter the electrical capacitance and inductance properties (bioreactance) of the chest. A 75-kHz current is applied and the relative phase shift of the transthoracic voltage is measured. The cardiac output (CO) is derived from the relative phase shift peak, HR, and ventricular ejection time. The following parameters from NICOM were studied: HR, MAP, SVV, stroke volume index (SVI), CI, and TPRI. The data were collected during the following time-points: before and at the end of study intervention, anesthetic induction (before and 1, 3, and 5 minutes after), tracheal intubation (before and 1, 3, and 5 minutes after), and positioning from supine to prone (before and 5 minutes after). We also collected data regarding baseline CAD using the ANSiscope (Dyansys Inc., Burlingame, CA, United States). The ANSiscope provides ANSindex (degree of dysautonomia), which is derived noninvasively from a recording of 572 RR intervals on an electrocardiogram.13

Conduct of Anesthesia

In the operating room, standard multiparameter monitors were attached. Anesthesia was induced with fentanyl 2 µg/kg, thiopentone 3 to 5 mg/kg, and preservative-free lignocaine 1 mg/kg, and vecuronium 0.15 mg/kg was used to facilitate intubation. Intubation was performed at one minimum alveolar concentration of sevoflurane administered in 50% oxygen-air mixture and anesthesia was maintained at the same level for the study period.

Hypotension during the study period was defined as a MAp < 60 mm Hg or a decrease in MAP by > 20% from the baseline. Hypotension was treated with mephentermine 3 mg boluses, repeated every 5 minutes until the MAP returned to the normal range. The incidence of hypotension, number of episodes of hypotension, total dose of mephentermine used, and total volume of intravenous fluids given during the study period were recorded.

Outcome Measures

Our outcome measures were changes in the HR, MAP, SVV, CI, and TPRI at predefined time points with study intervention (fluid loading), anesthetic induction, tracheal intubation, and prone positioning.

Sample Size Estimation and Statistical Analysis

The sample size was calculated for our primary objective (postinduction hypotension) based on a similar study where the incidence of hypotension was 52% versus 84% at 3 minutes postinduction in preloading versus no preloading groups.14 Considering an α error of 5 and 80% power, 30 patients in each group were deemed appropriate.

The data were analyzed using R software version 3.5.2. Data summaries are depicted as mean and standard deviation (SD) for interval scale variables, median and interquartile range (IQR) for ordinal variables, and frequencies and percentages for nominal variables. Odds ratio (OR) and 95% confidence interval (95% CI) are reported where applicable. Between-group differences were analyzed using Mann–Whitney U test for interval scale and ordinal data, while chi-square test was used for nominal data. Influence of group on trend of change of study variables within predefined time sets was analyzed using linear mixed-effects modeling. Unstructured covariance structure was assumed. For parsimony in analysis, only the first and last time points of each time set were used for modeling. We considered p < 0.05 as level for statistical significance.

Preloading and Hypotension in the Early Intraoperative Period

Hypotension after the induction of GA is a common occurrence. If severe and sustained, hypotension during the intraoperative period can lead to adverse consequences. Previous studies have documented benefits of fluid preloading in reducing the incidence of postinduction hypotension. A lower incidence (52 vs. 84%) of postinduction hypotension was observed with 20 mL/kg of Ringer's lactate preloading as compared with no preloading.14 Another study also documented lower incidence (12.5 vs. 57.5%) of postinduction hypotension with 6 mL/kg crystalloid loading as compared with control.15 Similarly, lower incidence (41.5 vs. 56.6%) of postinduction hemodynamic instability was seen with preinduction preloading of 8 mL/kg Ringer's acetate as compared with no preloading.7 In our study, too, we observed a significant difference in the occurrence of hypotension (26.7 vs. 86.7%) with and without fluid preloading. An earlier study demonstrated preoperative volume loading with 8 mL/kg Ringer's lactate to be more effective than preinduction administration of ephedrine in maintaining hemodynamic stability during rapid-sequence induction with propofol without HR increase during intubation.16 Our findings are in concurrence with the observations in previous studies. A comparatively higher incidence of hypotension is likely due to the presence of early CAD in our study population with CCCM. In contrast, a recent RCT comparing SV optimization by fluid loading versus standard fluid therapy before induction of anesthesia did not observe a significant difference in hemodynamic instability in the early intraoperative period.8

Cervical Myelopathy, Autonomic Dysfunction, and Hypotension

The cardiac autonomic system plays an important role in the occurrence of hemodynamic changes during anesthesia.17 The incidence of hypotension and vasopressor use after anesthetic induction for ophthalmic surgery was higher in diabetic patients with CAD compared with nondiabetics without CAD.18 The incidence of postinduction hypotension was 67 to 83% versus 9 to 17% in patients with and without CAD aged ≥40 years undergoing day-care surgery.19 An earlier study documented subclinical CAD in patients with CCCM.20 Patients in both the groups in our study had early CAD. These patients can have an exaggerated hemodynamic response to events such as induction of anesthesia, laryngoscopy and intubation, and prone positioning for surgery. The vulnerability for hypotension during the early intraoperative period increases further in patients with old age, male gender, higher ASA grade, lower MAP, emergency surgery, use of concomitant regional anesthesia, and larger dose of anesthetic induction agent.1 6 There was no difference between the groups for the above-mentioned risk factors in our study and hence the observed differences in postinduction hypotension are likely to be from the presence or absence of fluid preloading.

Effect of Prone Positioning on Hemodynamic and Cardiac Variables

Patients with CCCM develop hypotension during positioning from supine to prone for decompressive spine surgery due to decrease in TPRI.21 In contrast, decreases in SV and CI were responsible for decrease in BP after change from supine to prone position during lumbar spine surgery.22 In an observational study involving 20 patients, preloading with 20 mL/kg Ringer's lactate before positioning from supine to sitting prevented the occurrence of hypotension without significant changes in CO, SVR, SVV, CI, and SV.23 Though the decrease in MAP with prone positioning was more pronounced in patients without preloading in our study, this difference was not statistically significant. Similarly, there was no significant difference in cardiac indices between the groups with change to prone position.

Advanced Cardiac Indices and Anesthesia

An increase in plasma concentration of propofol within the therapeutic range resulted in decrease in MAP, central venous pressure, mean systemic filling pressure, venous return pressure, and SVR, increase in SVV, and no change in CO in 17 patients undergoing upper abdominal surgery.24 Thus, monitoring of CO and associated hemodynamic indices provides better insight into the probable mechanisms for hypotension after GA.

The SVV (%) at baseline was high (16.4 vs. 15.7) in both the groups in our study probably due to low intravascular volume from preoperative fasting. The SVV decreased significantly after fluid loading in the intervention group, while it did not change much in the control group. The SVV remained significantly lower at all the time points of the study period in preload group as compared with no preload group. The SVV decreased after anesthetic induction and increased after intubation and prone positioning in both the groups.

CI decreased from baseline after induction of anesthesia and after prone positioning in patients without preloading, though these changes were not significant. In patients who received preload, CI was maintained above the baseline values at all study time points. However, the difference between the groups was not statistically significant. The effect on CI reflected the combined effect of SVI and HR.

TPRI in both the groups decreased following induction of anesthesia, increased after intubation, and decreased with prone positioning. Preloading did not have any significant effect on the TPRI.

Strengths and Limitations of the Study

To the best of our knowledge, this is the first RCT comparing the effect of fluid preloading on the incidence of hypotension after the induction of GA along with dynamic cardiac indices during the early intraoperative period in patients with early CAD from CCCM. However, an important limitation of this study is restriction of study duration to the early intraoperative period. While the effect of fluid loading on hypotension and other cardiac indices is possible throughout the intraoperative period, other factors such as surgical nociception and blood loss are more likely to affect these outcomes than preanesthetic fluid loading. Second, we defined postinduction hypotension as MAP < 60 mm Hg or a decrease in MAP by >20% from the baseline for this study. The incidences are likely to be different with alternative definitions. Third, the volume of clear fluid intake during the preoperative period was not taken into account and therefore its contribution to postinduction hypotension in both the groups remains unknown. However, in the absence of difference in SVV (a measure of intravascular volume status) at baseline between the groups, influence of preoperative fluid intake is less likely. Fourth, the findings of our study are restricted to thiopentone anesthetic induction with sevoflurane maintenance. Lastly, being a single-center study involving patients with early CAD limits the generalizability of our findings to other populations.