Intraoperative changes in whole-blood viscosity in patients undergoing robot-assisted laparoscopic prostatectomy in the steep Trendelenburg position with pneumoperitoneum: a prospective nonrandomized observational cohort study

Background The aim of this study was to investigate the effect of the steep Trendelenburg position (STP) with pneumoperitoneum on whole-blood viscosity (WBV) in patients undergoing robot-assisted laparoscopic prostatectomy (RALP). The study also analyzed the associations of clinical patient-specific and time-dependent variables with WBV and recorded postoperative outcomes. Methods Fifty-eight adult male patients (ASA physical status of I or II) undergoing elective RALP were prospectively analyzed in this study. WBV was intraoperatively measured three times: at the beginning of surgery in the supine position without pneumoperitoneum; after 30 min in the STP with pneumoperitoneum; and at the end of surgery in the supine position without pneumoperitoneum. The WBV at a high shear rate (300 s− 1) was recorded as systolic blood viscosity (SBV) and that at a low shear rate (5 s− 1) was recorded as diastolic blood viscosity (DBV). Systolic blood hyperviscosity was defined as > 13.0 cP at 300 s− 1 and diastolic blood hyperviscosity was defined as > 4.1 cP at 5 s− 1. Results The WBV and incidences of systolic and diastolic blood hyperviscosity significantly increased from the supine position without pneumoperitoneum to the STP with pneumoperitoneum. When RALP was performed in the STP with pneumoperitoneum, 12 patients (27.3%) who had normal SBV at the beginning of surgery and 11 patients (26.8%) who had normal DBV at the beginning of surgery developed new systolic and diastolic blood hyperviscosity, respectively. The degree of increase in WBV after positioning with the STP and pneumoperitoneum was higher in the patients with hyperviscosity than in those without hyperviscosity at the beginning of surgery. Higher preoperative body mass index (BMI) and hematocrit level were associated with the development of both systolic and diastolic blood hyperviscosity in the STP with pneumoperitoneum. All patients were postoperatively discharged without fatal complications. Conclusions Changes in surgical position may influence WBV, and higher preoperative BMI and hematocrit level are independent factors associated with the risk of hyperviscosity during RALP in the STP with pneumoperitoneum. Trial registration Clinical Research Information Service, Republic of Korea, approval number: KCT0003295 on October 25, 2018.


Background
Prostate cancer is one of the most common cancers in the male population, and its incidence gradually increases with age [1]. Based on the characteristics of this cancer, surgical resection is an appropriate option to improve patient outcome. In particular, robot-assisted laparoscopic prostatectomy (RALP) is a technically advanced surgical method that has been widely accepted as feasible and effective due to its various benefits, including its minimally invasive nature, improved prognosis, and favorable functional results [2,3]. Additionally, RALP has advantages compared to conventional prostatectomy, including higher quality stereoscopic surgical view and increased maneuverability [4]. However, patients who undergo RALP are typically placed in a specific surgical position-the steep Trendelenburg position (STP)-with pneumoperitoneum using CO 2 gas. This surgical position may be associated with the development of complications, such as subcutaneous emphysema, pulmonary atelectasis, and increased airway and/or optic pressure [5]. Eventually, intraoperative development of these pathophysiological events may become challenging to both urologists and anesthesiologists during RALP.
Whole-blood viscosity (WBV) plays an important role in circulatory flow in both large and small vessels [6]. Because whole blood exhibits non-Newtonian shearthinning viscosity behavior, such that the velocity of circulatory flow is inversely related to the degree of WBV, the WBV is often regarded as a functional marker of shear stress within vessels [7]. This viscosity is also associated with features of blood components, such as deformability, aggregation, and concentration of red blood cells and other proteins [6,8,9]. Because increased WBV may cause greater oscillation of shear stress in relation to endothelial injury, close association of WBV with the development of cardio-and/or cerebrovascular diseases has been reported in clinical settings [10][11][12][13][14]. Additionally, various risk factors, such as obesity, aging, and comorbidity, seem to contribute to increases in WBV [6,9,[15][16][17][18]. Because the age of patients undergoing RALP progressively increases with the prevalence of various comorbidities, such as diabetes mellitus (DM) and hypertension [19], these patients may be vulnerable to the development of hyperviscosity during surgery.
The aim of this study was to investigate the effect of the STP with pneumoperitoneum on WBV in patients undergoing RALP. The study also analyzed the associations of clinical variables with WBV and recorded postoperative outcomes.

Ethical considerations
This single-center, prospective, nonrandomized observational cohort study was approved by the Institutional Review Board of Seoul St. Mary's Hospital Ethics Committee (approval number: KC18OESI0540). The study protocol was retrospectively/prospectively registered at a publicly accessible clinical registration site that is recognized by the International Committee of Medical Journals Editors (Clinical Research Information Service, Republic of Korea, approval number: KCT0003295, https://cris.nih.go.kr/cris/en/search/ search_result_st01.jsp?seq=13745). Written informed consent was obtained from all patients at our hospital who were enrolled between September 2018 and February 2019.

Study population
Inclusion criteria for this study were male gender, age ≥ 19 years, scheduled for elective RALP, and an American Society of Anesthesiologists (ASA) physical status of I or II [20]. Exclusion criteria were emergency case, age < 19 years, ASA physical status of III-V [20], intraoperative development of hemodynamic instability that required rescue management, such as colloid infusion, blood product transfusion, or strong vasopressor administration (i.e., epinephrine or norepinephrine), and refusal to participate in the study, because patients with symptomatic or uncontrolled diseases may have various confounders related to changing WBV (i.e., possibly, medications and/or nature of own disease). A total of 60 patients agreed to participate in our study; however, 2 patients experienced massive hemorrhage during surgery that required colloid infusion and blood product transfusion, and were thus excluded from the analysis. Therefore, 58 male patients were enrolled in this study.

Anesthesia and surgery
Balanced anesthesia was performed by attending expert anesthesiologists. Induction of anesthesia was achieved using 1-2 mg.kg − 1 propofol (Fresenius Kabi, Bad Homburg, Germany) and 0.6 mg.kg − 1 rocuronium (Merck Sharp & Dohme Corp., Kenilworth, NJ, USA); anesthesia was then maintained using 2.0-6.0% desflurane (Baxter, Deerfield, IL, USA) under medical air in oxygen. Remifentanil (Hanlim Pharm. Co., Ltd., Seoul, Republic of Korea) was continuously infused at a rate of 0.1-0.5 μg.kg − 1 .min − 1 , as appropriate. The Bispectral Index™ instrument (Medtronic, Minneapolis, MN, USA) was set between 40 and 50 to ensure appropriate hypnotic depth. Rocuronium was repeatedly infused under trainof-four monitoring (> one twitch). End-tidal CO 2 was set between 30 and 40 mmHg with adjustment of the ventilator mode. Central venous pressure (CVP) was monitored using a central venous catheter (Arrow, Morrisville, NC, USA) that was inserted by experienced radiologists on the day before surgery. For fluid therapy [21], a baseline isotonic crystalloid was prepared based on the estimated fluid maintenance requirements, in turn based on the patient's weight and anticipated tissue trauma. Additional fluid boluses were infused according to blood loss; however, the total amount of fluid was restricted to a maximum of 1 L before vesicourethral anastomosis.
Surgery was exclusively performed by expert urologists using a specific robot-assisted laparoscopic method. CO 2 gas was insufflated into the abdominal cavity (i.e., pneumoperitoneum) in the supine position, and patients were placed in the STP with the maximal angle (45°) of the surgical table (Maquet, Rastatt, Germany); this approach was routinely applied to achieve the optimal surgical view. Intra-abdominal pressure was maintained at 12-15 mmHg during surgery. At the time of peritoneal closure, the surgical position was restored to the supine position with removal of CO 2 gas.

Measurement of WBV
WBV was measured during surgery three times via a central venous catheter: at the beginning of surgery (i.e., skin incision) in the supine position without pneumoperitoneum, after 30 min in the STP with pneumoperitoneum, and at the end of surgery (i.e., peritoneal closure) in the supine position without pneumoperitoneum (Table 1). Blood samples (3 mL) were collected into evacuated test tubes (BD Vacutainer, K2 EDTA; Becton, Dickinson and Company, Franklin Lakes, NJ, USA) without venous stasis, and measured using an automated scanning capillary tube viscometer (Hemovister; Ubiosis, Seongnam, Republic of Korea): WBV at a high shear rate (300 s − 1 ) was recorded as systolic blood viscosity (SBV) and that at a low shear rate (5 s − 1 ) was recorded as diastolic blood viscosity (DBV) [22]. The reference intervals for WBV were 3.5-4.1 cP at 5 s − 1 and 9.4-13.0 cP at 300 s − 1 in male patients, based on the manufacturer's instructions. Accordingly, systolic blood hyperviscosity was defined as > 13.0 cP at 300 s − 1 and diastolic blood hyperviscosity was defined as > 4.1 cP at 5 s − 1 .

Statistical analysis
Because of the exploratory nature of this study and the lack of a primary endpoint with anticipated effect sizes, no sample size or statistical power calculations were performed. To achieve robust effects within a reasonable interval of time, we chose to collect data from approximately 60 patients, considering the possibility of patient drop-out. The sample size was based on available data from all patients who underwent elective RALP at our hospital during a 1-year period. A post hoc power analysis revealed that a sample size of 58 was required for an expected power of 99%. The effect size r (r = z/sqrt(n)) was 0.82 (|-6.234|/ sqrt(58)) and 0.86 (|-6.569|/sqrt(58)), with respect to SBV and DBV. The statistical analysis was calculated using the Wilcoxon signed-rank test (one sample case), as a post hoc method, included in G*Power software 3.1.9.4 (http://www.gpower.hhu.de/). The normality of the distribution of continuous data was evaluated using the Shapiro-Wilk test. Continuous data are expressed as median and interquartile range (IQR), while categorical data are expressed as number and proportion. Serial changes in WBV and vital signs were analyzed using the Friedman's test and the Wilcoxon signed-rank test with the Bonferroni's post hoc method. We compared the levels of WBV and vital signs between the beginning of surgery (reference) vs. the STP with pneumoperitoneum and the end of surgery (p < 0.025 was statistically significant). After separation of WBV into normal viscosity and hyperviscosity, intraoperative changes in the proportions of patients with hyperviscosity were analyzed using Cochran's Q test with the McNemar post hoc  test. Incidence rates of systolic and diastolic blood hyperviscosity according to surgical position were compared using the χ 2 test or Fisher's exact test, as appropriate.
The associations between clinical variables and the development of whole-blood hyperviscosity during RALP in STP with pneumoperitoneum were evaluated using univariate and multivariate logistic regression analyses. Potentially significant factors (p < 0.1) in univariate analyses were entered into multivariate forward and backward regression analyses. The values are expressed as odds ratios and 95% confidence intervals (CIs). The clinically relevant factors were chosen when multiple perioperative factors were inter-correlated. All categorical variables were modeled as dummy variables. The predictive power of independent factors for hyperviscosity were assessed using the area under the receiver operating characteristic curve (AUC). The optimal preoperative BMI and hematocrit cut-off values according to hyperviscosity at the STP with a pneumoperitoneum were determined using the AUC method. All tests were twosided, and p < 0.05 was considered statistically significant. Statistical analyses were performed using SPSS (ver. 22.0 for Windows; SPSS Inc., Chicago, IL, USA) and MedCalc (ver. 11.0 for Windows; MedCalc Software, Mariakerke, Belgium). We used the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) criteria for our observational study [25].

Preoperative and intraoperative findings in patients undergoing RALP
In the present study, all patients had ASA physical status I or II and underwent elective RALP.   (Table 3). With respect to vital signs, SBP, DBP, and CVP significantly increased, while HR decreased, from the supine position without pneumoperitoneum (beginning of surgery) to the STP with pneumoperitoneum. There was no difference in vital signs between the beginning and end of surgery when both were assessed in the supine position without pneumoperitoneum.

Intraoperative changes in SBV and DBV
Levels of SBV and DBV significantly increased from the supine position without pneumoperitoneum (beginning of surgery) to the STP with pneumoperitoneum ( Table 4). The incidence rates of systolic and diastolic blood hyperviscosity were higher in the STP with pneumoperitoneum than in the supine position without pneumoperitoneum (beginning of surgery). However, there were no differences in the levels of SBV and DBV and incidence rates of systolic and diastolic blood hyperviscosity between the beginning and end of surgery when both were assessed in the supine position without pneumoperitoneum.
With respect to changes in the prevalence rates of systolic and diastolic blood hyperviscosity during surgery (Table 5), 12 patients (27.3%) who had normal SBV in the supine position without pneumoperitoneum (beginning of surgery) showed new-onset systolic blood hyperviscosity in the STP with pneumoperitoneum. In addition, 11 patients (26.8%) who had normal DBV in the supine position without pneumoperitoneum (beginning of surgery) showed new-onset diastolic blood hyperviscosity in the STP with pneumoperitoneum. All patients with systolic (n = 14) or diastolic (n = 17) blood hyperviscosity in the supine position without pneumoperitoneum (beginning of surgery) continued to show hyperviscosity in the STP with pneumoperitoneum. In patients with persistent hyperviscosity, the changes in SBV (n = 14) were as follows There were significant differences in SBV (p = 0.001) and DBV (p < 0.001) between the supine position without pneumoperitoneum (beginning of surgery) and the STP with pneumoperitoneum.
WBV significantly increased from the supine position without pneumoperitoneum to the STP with pneumoperitoneum in patients with/without hyperviscosity at the beginning of surgery (i.e., supine position without pneumoperitoneum) (Additional files 1 and 2); however, the degree of increase in WBV after positioning with the STP and pneumoperitoneum was higher in patients with hyperviscosity than in those without hyperviscosity.
Association of the development of systolic and diastolic blood hyperviscosity with clinical factors in patients undergoing RALP in the STP with pneumoperitoneum With respect to systolic blood hyperviscosity in the STP with pneumoperitoneum (Table 6), preoperative findings (i.e., age, BMI, hypertension, hematocrit level, and blood urea nitrogen level) and one intraoperative finding (i.e., CVP in the STP with pneumoperitoneum) were potentially associated with the development of hyperviscosity in univariate analysis. Higher preoperative BMI and hematocrit level were significantly associated with the development of hyperviscosity in multivariate analysis. The predictive accuracy of this model was good (AUC = 0.821; 95% CI = 0.708-0.934; p < 0.001). The optimal cut-off values for preoperative BMI and hematocrit were  (Table 7), preoperative findings (i.e., age, BMI, and hematocrit level) and intraoperative findings (i.e., DBP and HR in the supine position without pneumoperitoneum [beginning of surgery]) were potentially associated with the development of hyperviscosity in univariate analysis. Higher preoperative BMI and hematocrit level were significantly associated with the development of hyperviscosity in multivariate analysis. The predictive accuracy of this model was fair (AUC = 0.790; 95% CI = 0.673-0.908; p < 0.001). The optimal cut-off values for preoperative BMI and hematocrit were 25.1 kg.m − 2 (AUC = 0.645; 95% CI = 0.509-0.766; p = 0.046) and 43% (AUC = 0.712; 95% CI = 0.579-0.824; p = 0.002), respectively, for hyperviscosity.

Observed complications
Nearly all patients in our study were discharged on postoperative day 6 (mean ± standard deviation: 6.2 ± 0.8 days) and there were no fatal complications representative of Clavien-Dindo class I [24].

Discussion
The main finding of this study was that performance of RALP in the STP with pneumoperitoneum may contribute to a significant increase in WBV (in either SBV or DBV), which subsequently results in the development of hyperviscosity. Additionally, the surgical position change with gas insufflation may have aggravated WBV more in patients with hyperviscosity at the beginning of surgery than in those without hyperviscosity. Higher preoperative BMI and hematocrit level were significantly associated with increases in both SBV and DBV when RALP was performed in the STP with pneumoperitoneum.
WBV has been linked to complicated interactions among blood components, such as red blood cells, fibrinogen, and lipid profiles. Moreover, WBV is separated into SBV and DBV, based on shear rates: SBV exhibits a high shear rate (≥ 300 s − 1 ) and DBV exhibits a low shear rate (≤ 5 s − 1 ) [14,26,27]. SBV is related to the frictional characteristics of rapid blood flow in larger vessels, which are affected by inertia force, rather than viscous force [14,26]. Notably, hematocrit and erythrocyte deformability mainly contribute to increased SBV [28]. In contrast, DBV is related to the frictional characteristics of low blood flow in smaller vessels (i.e., distal or peripheral vascular segments), which are affected by viscous force, rather than inertia force; notably, DBV predominantly determines tissue perfusion [14]. Plasma proteins and lipids, such as fibrinogen and LDL-C, largely contribute to increased DBV, together with hematocrit and red blood cell aggregation [29].
Many clinical studies have suggested that elevated WBV is associated with the development of cardiovascular and cerebrovascular diseases [10,11,13,30]. In patients with mitral annular calcification, higher WBV was predictive of pathologic mitral annulus. Notably, viscosity was inversely correlated with early and late diastolic mitral annular velocity, which are represented by mitral annular limitations [13]. In patients with acute ischemic stroke, increased DBV was significantly associated with the prevalence of stroke after adjustment for various vascular risk factors, such as age and sex; viscosity was also correlated with hematocrit level at admission. Notably, viscosity decreased after hydration therapy, but hyperviscosity recurred when hydration therapy was discontinued [11]. Our study population showed a more stable clinical condition before surgery, as represented by ASA Table 4 Changes in systolic and diastolic blood viscosities during robot-assisted laparoscopic prostatectomy physical status I or II, compared to the patients in previous studies [10,11,13,20,30]. During RALP surgery, we encountered no hemodynamic instability that required aggressive fluid or blood product transfusion therapy. However, the change in surgical condition (i.e., to STP with pneumoperitoneum) led to dramatic increases in both intra-abdominal/thoracic pressure (i.e., CVP) and WBV (i.e., SBV and DBV); and caused more stickiness of WBV in the patients with hyperviscosity at the beginning of surgery than in those without hyperviscosity. Therefore, establishing intraoperative STP with a pneumoperitoneum may contribute to changes in WBV. Subsequently, when patients were returned to the supine position without pneumoperitoneum, both intra-abdominal/ thoracic pressure and WBV were reduced. Although there was a lack of clinical endpoints (i.e., morbidity and/or mortality) in our study, our results demonstrate a relationship between intraoperative WBV and surgical STP position with gas insufflation, which is usually present during laparoscopy-based surgery [5], and the impact of surgical position with gas insufflation on aggravation of the WBV condition in a relatively healthy population (i.e., ASA physical status of I or II). Although further study is required, these findings imply that patients with hyperviscosity are more susceptible to aggravated WBV than those without hyperviscosity when the STP with pneumoperitoneum is performed during surgery. The mechanism by which the WBV is increased in the STP with pneumoperitoneum remains unclear. In the present study, higher BMI and hematocrit level before surgery were significantly related with increases in both SBV and DBV in the STP with pneumoperitoneum. First, obesity has been linked to increases in hemorheological parameters, including WBV [18,31]. In obese patients without comorbidities, significant increases in blood viscosity have been observed, which may be related to poor erythrocyte aggregation and deformability [9]. In a population of female volunteers, blood viscosity was positively correlated with overall body fat status, as measured by BMI; in addition, increased BMI was correlated with increased red blood cell rigidity [32]. Our study findings were consistent with those of previous studies [9,18,31,32], in that elevated BMI was significantly associated with increased WBV. However, our findings differed from those of previous studies [9,18,31,32]: WBV in previous study populations was evaluated in the resting supine position, whereas in our study, WBV was serially evaluated according to changes in surgical position with/without pneumoperitoneum, which may affect systemic circulatory pressure or flow [33][34][35]. Therefore, increased WBV was more pronounced in the STP with pneumoperitoneum than in the supine position without pneumoperitoneum; this finding may be due to increasing intra-abdominal/thoracic pressure, which led to the compression of vessels related to impairment of circulatory flow [17,33,36,37]. Compared to non-obese patients, obese patients in the STP with pneumoperitoneum showed increased pressure of abdominal contents on cardiovascular circulatory flow more frequently [38]. Although further studies are needed to identify the potential cascade underlying the development of hyperviscosity, higher BMI may have a greater impact on increased blood viscosity during RALP in the STP with pneumoperitoneum.
Hematocrit is an important factor that is positively related to WBV [8]. In the clinical setting, increases in WBV and hematocrit level have been associated with the risk of cardiovascular disease. In particular, small vessel flow and perfusion, such as coronary  collateral circulation, is most affected by elevated WBV [12]. Because high WBV through wall shear stress seems to be directly related to the rupture or erosion of weak plaques, therapeutic phlebotomy to decrease hematocrit may select for patients at risk of atherosclerotic occlusions. Additionally, monitoring of WBV was useful for selection of appropriate time points to perform phlebotomy, which could reduce hematocrit level and, subsequently, WBV [39]. However, previously, the associations among hematocrit, WBV, and surgical position were not investigated. In the operating theater, patients are required to assume various positions that are not physiologically normal, such as the STP. Furthermore, consistent with advanced surgical methods, including robot-assisted laparoscopic surgery, CO 2 gas insufflation in the abdominal cavity (i.e., pneumoperitoneum) is used to achieve an optimal surgical view, but this may impair hemodynamic flow [33,35,36]. In the present study, the relationship between hematocrit and WBV may have been strengthened in the STP with pneumoperitoneum, compared to the supine position without pneumoperitoneum. This hypothesis was supported by the observation that more than one-quarter of patients with normal viscosity showed development of hyperviscosity after changing position, from the supine position without pneumoperitoneum to the STP with pneumoperitoneum. At the end of surgery (i.e., the supine position without pneumoperitoneum), the WBV recovered to the level observed at the beginning of surgery (i.e., the supine position without pneumoperitoneum). Additionally, in patients who showed hyperviscosity at the beginning of surgery (i.e., the supine position without pneumoperitoneum), the WBV further increased in the STP with pneumoperitoneum, and then decreased at the end of surgery (i.e., the supine position without pneumoperitoneum).
There were some limitations to this study. First, because of logistical restrictions, we did not evaluate the WBV prior to surgery. Neither hemodynamic instability nor resuscitation therapy were observed before or after anesthetic induction in our cohort. Second, we did not measure the deformability of red blood cells. Because red blood cell deformability affects WBV [28,29], further analyses may be helpful to identify potential causes of increased WBV during surgery in the STP with pneumoperitoneum, and to subsequently restore normal viscosity in the supine position without pneumoperitoneum. Third, we excluded patients who had uncontrolled comorbidities (ASA physical status III or IV), such as DM and hypertension [20]. All patients were discharged without fatal complications (i.e., Clavien-Dindo class I) [24]. Our exclusion criteria may have precluded an effect of comorbidities on WBV and postoperative outcomes [40]. Therefore, an additional large-scale cohort study of patients undergoing RALP is needed to evaluate the association between STP and WBV, and to make suggestions with respect to the management of increased WBV during surgery.

Conclusions
We found that WBV (i.e., SBV and DBV) significantly increased, in a time-dependent manner, in patients undergoing RALP in the STP with pneumoperitoneum. The WBV reached levels indicative of hyperviscosity, but no patient showed occlusions of hemodynamic circulation, which can result from the recovery of normal WBV at the end of surgery (i.e., the supine position without pneumoperitoneum). The patients with hyperviscosity at  the beginning of surgery (i.e., the supine position without pneumoperitoneum) seemed to be more susceptible to aggravation of WBV than those without hyperviscosity when the STP with pneumoperitoneum was performed during surgery. Therefore, changes in surgical position may contribute to changes in WBV. Higher preoperative BMI and hematocrit level were significantly associated with the development of hyperviscosity during RALP in the STP with pneumoperitoneum. An additional large-scale study may be helpful for identification of patients at risk of hyperviscosity and circulatory impairment during surgery requiring use of the STP with pneumoperitoneum.