High Central Venous Pressure and Impaired Microcirculation During Pneumoperitoneum: Implications for Perioperative and Obstetric Anesthesia

VJM Spring Edition 2026

https://doi.org/10.18130/753k-tx80

Author: Fady E. Attia, BS¹ Nabil Elkassabany MD²

Author Affiliation: 

¹ University of Virginia School of Medicine, Charlottesville, Virginia

² Department of Anesthesiology, University of Virginia, Charlottesville, Virginia

Conflicts of Interest:

None declared.

---

Abstract:

Esophagectomy is known to be accompanied by a high rate of perioperative morbidity and mortality. This is, at least in part, because gastric conduit perfusion is sensitive to the conditions of minimally invasive surgery with pneumoperitoneum and reverse-Trendelenburg positioning. The impact of these procedures on venous return and microcirculatory flow under goal-directed hemodynamic conditions is not well understood. This prospective observational pilot study was conducted on eleven  adult patients undergoing minimally invasive esophagectomy. Hemodynamic conditions were managed according to a goal-directed fluid optimization strategy. Cardiac output and mean systemic filling pressure (MSFP) were calculated from arterial waveform analysis, and sublingual microcirculatory flow was evaluated by sidestream dark-field imaging with visual microvascular flow index (MFI) scoring. Measurements were obtained at three intraoperative time points: baseline supine, pneumoperitoneum supine, and pneumoperitoneum with 20° reverse‑Trendelenburg. Pneumoperitoneum and reverse‑Trendelenburg significantly increased central venous pressure (CVP; 16±5 to 21±5 mmHg), MSFP (22±4 to 27±5 mmHg), and stroke volume variation, while reducing MFI (2.1±0.5 to 1.7±0.4; all P≤0.031). In contrast, cardiac output, mean arterial pressure, pressure gradient for venous return, and resistance to venous return did not change significantly across time points. Higher CVP and MSFP were associated with lower MFI, while MFI was not significantly related to cardiac output or arterial pressure. In pneumoperitoneum with reverse-Trendelenburg positioning, venous congestion was linked to poor sublingual microcirculatory flow despite normal macrocirculatory parameters. For anesthesiologists, these findings emphasize the significance of avoiding high venous pressures rather than focusing solely on blood pressure and cardiac output.

---

Introduction

Esophagectomy remains a high-risk procedure with morbidity approaching 40% and substantial mortality, largely due to anastomotic complications in the gastric conduit. The gastric conduit's vulnerability to ischemia makes maintaining tissue perfusion a key priority for anesthesiologists and surgeons during minimally invasive esophagectomy.¹

For optimal visualization during thoracolaparoscopic esophagectomy, pneumoperitoneum and steep reverse-Trendelenburg positioning are used, and both maneuvers are known to have a significant effect on venous return and cardiac function. The increase in intra-abdominal pressure can compromise tissue perfusion even when normal physiological parameters are maintained. Previous studies have demonstrated that increased central venous pressure can have a negative effect on microvascular flow and clinical outcomes in critical illness, although there are no specific data for complex upper GI surgery.^2–4

This prospective observational pilot study aims to evaluate the effect of pneumoperitoneum and steep reverse-Trendelenburg positioning on mean systemic filling pressure, venous return, and sublingual microcirculation during minimally invasive esophagectomy. By integrating continuous macrocirculatory monitoring with direct microcirculatory imaging, the investigators sought to clarify whether restoring cardiac output and blood pressure is sufficient to ensure adequate microcirculatory perfusion, or whether elevated venous pressures themselves may be harmful.

The results of this study may be beneficial to anesthesiologists, especially obstetric anesthesiologists who are often faced with high venous pressures, volume status issues, and intricate physiological principles. The purpose of the Clinical Trial Summary is to provide an overview of the design of the study, provide quantitative results, and highlight the implications for anesthesia care, including the equity issue in obstetric anesthesia.

Study Design and Methods

The study was a prospective observational study and was conducted in a single teaching hospital, directly in the operating room. The study recruited eleven consecutive adult patients undergoing minimally invasive esophagectomy, including transthoracic and transhiatal approaches with cervical and intrathoracic anastomoses. Exclusion criteria were preexisting severe heart conditions such as left ventricular ejection fraction less than 40%, valvular heart disease, and known allergies to colloids. The study was approved by the ethical committee, and informed consent was waived.

Anesthetic technique and perioperative management

For thoracic epidural analgesia, the procedure was placed at T5–6 or T6–7 if feasible. For general anesthesia, propofol (about 2–3 mg/kg) was used for induction, together with sufentanil and rocuronium (1.2 mg/kg) to facilitate intubation, and was maintained with sevoflurane at one minimum alveolar concentration. Invasive monitoring was done with radial arterial line and right subclavian central line, together with pulse oximetry and capnography. For epidural analgesia, intermittent boluses of bupivacaine with sufentanil, followed by continuous infusion of bupivacaine, was used. If epidural placement was not feasible, multimodal systemic analgesia with ketamine and/or lidocaine infusion was used.

For hemodynamic management, fluid optimization was done with pulse contour analysis (EV1000, Edwards Lifesciences) to optimize stroke volume (SV). After induction and stabilization of mean arterial pressure (MAP) to or above 65 mmHg, or within 20% of baseline, repeated 250 mL colloid boluses (tetraspan 6%) were given to increase SV until it stopped increasing by more than 10%, which was considered “optimal.” Crystalloids were given at 1 mL/kg/h, and colloid boluses were given when SV was more than 10% below the target. Vasopressors, such as norepinephrine, were titrated to achieve MAP goals, and heart rate was kept less than 100 beats per minute if feasible.

Surgical approach and intra‑abdominal pressure

The thoracolaparoscopic esophagectomies with gastric tube reconstructions were performed on all patients. Insufflation of CO2 into the abdomen reached a peak intra-abdominal pressure of 15 mmHg using a supraumbilical port. Patients were then placed in an approximately 20 degrees reverse-Trendelenburg position from the supine position after placement of the trocars. A peak intrathoracic pressure of 6 mmHg was maintained during the thoracoscopic phase.

Hemodynamic and microcirculatory measurements

The global hemodynamics were measured and tracked continuously, including systolic and diastolic pressures, mean arterial pressure, central venous pressure, cardiac output, heart rate, stroke volume, and stroke volume variation. The analogue model calculated the MSFP with the help of CVP, MAP, CO, and patient size, considering the arterial and venous compartments to be compliant and the resistive pathways. The pressure gradient for venous return (PVR) and the resistance to venous return (Resistance-vr) were calculated with the help of the aforementioned parameters: PVR = MSFP - CVP and Resistance-vr = [MSFP - CVP]/CO.

The sublingual microcirculation was studied with the help of sidestream dark-field video microscopy, and three areas were studied at each time point. Each video recording lasted for about 20 seconds and was checked for image quality, sharpness, and absence of pressure artifacts. The flow in the sublingual microvessels was rated with the help of a score called the microvascular flow index (MFI), ranging from 0 to 3, where 0 means no flow, 1 means intermittent flow, 2 means sluggish flow, and 3 means continuous flow. An MFI score less than 2 indicates a severely compromised state of microvascular flow.​

Measurement time points and statistical analysis

Hemodynamic and microcirculatory data were recorded at three specific time points: at baseline and in a supine position before pneumoperitoneum (T1), after pneumoperitoneum was established in a supine position (T2), and after a pneumoperitoneum setup with a steep reverse-Trendelenburg position (T3). Each data point was recorded over a time period of approximately five minutes from the respective time point under conditions that were stable according to our fluid and vasopressor therapy protocol.

Results are described as continuous data with means and standard deviation and as categorical data with absolute and relative frequencies. To evaluate time-related changes from T1 to T2 and from T2 to T3, a General Linear Model for Repeated Measures was applied, which corrects for data from the same subjects. To evaluate specific data points, a paired test (t-test or Wilcoxon test as appropriate) was applied. Associations between MFI data were evaluated with Spearman correlation tests. To evaluate which thresholds for macrohemodynamic parameters predict low microcirculatory flow index (<2), ROC curve analyses were applied with P values <0.05 considered statistically significant.

Key Results

Patient characteristics and intraoperative management

The study recruited eleven patients with a mean age of 64.0 ± 9.5 years, and their gender distribution was one female and ten males. The history of hypertension was present in 27% (3/11) and type 2 diabetes mellitus in 9% (1/11) of patients. During the procedure, 91% (10/11) required norepinephrine to support their blood pressure during anesthesia and pneumoperitoneum. The types of surgeries were five transhiatal, four transthoracic with cervical anastomosis, and one transthoracic with intrathoracic anastomosis. The mean time for the procedure was 337 ± 141 minutes, and the median blood loss was 150 mL (interquartile range 50–400 mL).

Evolution of macrocirculation and microcirculation across time points

Pneumoperitoneum and reverse-Trendelenburg significantly affected venous pressures and stroke colume variation (SVV), while global circulation flow and arterial blood pressure remained relatively stable. Central venous pressure increased significantly from 16 ± 5 mmHg at T1, to 20 ± 5 mmHg at T2, to 21 ± 5 mmHg at T3 (P < 0.0001 for repeated measures analysis). Mean systemic filling pressure increased from 22 ± 4 mmHg at T1 to 25 ± 5 mmHg at T2 to 27 ± 5 mmHg at T3 (P = 0.001). SVV increased significantly from 9 ± 3% at T1 to 13 ± 7% with pneumoperitoneum and 19 ± 10% with pneumoperitoneum and reverse-Trendelenburg (P = 0.005).

Cardiac output remained relatively stable throughout the time points (4.3 ± 0.9 L/min/m² at T1, 4.3 ± 1.3 L/min/m² at T2, 4.4 ± 1.6 L/min/m² at T3; P = 0.110). Mean arterial blood pressure did not differ significantly between time points (73 ± 8 mmHg at T1, 82 ± 17 mmHg at T2, 85 ± 18 mmHg at T3; P = 0.956). Pressure gradient for venous return and resistance to venous return did not differ significantly between time points.

Microcirculatory flow in the sublingual vessels deteriorated with time; the microvascular flow index decreased significantly from 2.1 ± 0.5 at T1 to 2.1 ± 0.7 with pneumoperitoneum and 1.7 ± 0.4 with pneumoperitoneum and reverse-Trendelenburg (P = 0.031). Perfused vessel density did not differ significantly. Direct comparisons between T1 and T3 showed that CVP and MSFP increased significantly with reverse-Trendelenburg position.​

Relationship between venous pressures and microcirculation

Pooling all time points, there were 33 paired macro- and microcirculatory data points. MFI was significantly negatively correlated with central venous pressure (CVP) (Spearman correlation r = -0.380, P = 0.029) and with MSFP (r = -0.367, P = 0.0367), showing that increased venous pressures were correlated with reduced sublingual microcirculatory flow. No statistically significant correlations were found between mean flow index and cardiac output (CO) (r = -0.024, P = 0.893) or between mean flow index and mean arterial pressure (MAP) (r = -0.254, P = 0.154).

When stratifying according to microcirculatory status, there were 13 data points with MFI < 2 and 20 data points with MFI ≥ 2. The low MFI group had increased venous pressure: central venous pressure was 21.8 ± 4.9 mmHg compared with 17.4 ± 4.8 mmHg in the high MFI group (P = 0.009), and mean systemic filling pressure was 28 ± 6 mmHg compared with 23 ± 5 mmHg (P = 0.013).​

Predicting severely impaired microcirculatory flow

ROC analysis assessed which macrocirculatory variables best discriminated MFI < 2. The area under the ROC curve (AUC) for CVP was 0.767 (95% confidence interval 0.588–0.896), with an optimal cutoff of 23 mmHg yielding 100% specificity and 61.5% sensitivity for predicting MFI < 2. MSFP had an AUC of 0.758 (95% confidence interval 0.578–0.889) with a cutoff of 29 mmHg, 95% specificity, and 61.5% sensitivity. In contrast, MAP and CO had more modest discriminative ability, with AUCs of 0.629 and 0.560, respectively.

Together, these findings indicate that during pneumoperitoneum and reverse‑Trendelenburg, venous congestion (as captured by elevated CVP and MSFP) was closely linked to impaired sublingual microcirculatory flow, even when global blood pressure and cardiac output were within target ranges.

Clinical and Equity Implications

This pilot study suggests to anesthesiologists that traditional “normal” limits of mean arterial pressure and cardiac output do not guarantee optimal tissue perfusion during complex and high-risk abdominal surgery with pneumoperitoneum. In well-volume-resuscitated patients, there was a significant increase in venous pressure with no change in cardiac output and mean arterial pressure, yet tissue perfusion was impaired. This suggests that increased tissue and vascular flows with aggressive fluid resuscitation and vasopressor use to increase central venous pressure can, in fact, impair microvascular flows.

CVP as a clinical instrument: CVP is classically considered to be an important indicator of preload. In this study, it is advanced as a putative predictor of impaired microvascular flows. It is apparent from these studies that when the CVP is 23 mmHg, it is very specific for low tissue flows. Perhaps it is time to rethink the “normal” limits of venous pressure when making clinical decisions.^5–7

These studies provide mechanistic explanations for clinical concerns and provide direct applicability to obstetric anesthesia practice. In pregnant patients, it is necessary to take into account increased circulating blood volume, changes in venous capacitance, and often comorbidities such as obesity and hypertension. This is because pregnant patients are particularly vulnerable to such exacerbations due to pregnancy’s unique hemodynamic profile, which predisposes to elevated venous pressures and reduced tissue perfusion.⁸ In these patient groups, increased CVP can be further exacerbated by pneumoperitoneum during non-obstetric laparoscopy, high airway pressures, and fluid resuscitation, which can cause venous congestion to the uterus, placenta, liver, and kidney, despite “normal” mean arterial pressures.

Limitations

However, there are a number of limitations to this study, which must be taken into consideration when drawing conclusions about causality or making broader applications. Firstly, the sample size is small, consisting of only eleven patients, providing 33 data pairs. This limits the degree of precision that can be attained, as well as increases the probability of type I and II error. It should also be noted that this study, performed within a highly advanced academic center with a strict goal-directed fluid management strategy, may not be applicable to other institutions or different surgical techniques.

The second limitation of this study is the inability to establish causality with regard to increased central venous pressure or mean systemic filling pressure and microcirculatory flow. Due to the interrelated natures of fluid management, vasopressor use, and length of pneumoperitoneum, there are a number of confounding factors that could have played a role, including changes in cardiac function, which could potentially have an impact on both central venous pressure as well as microvascular flow index. Notably, cardiac function was not directly assessed with echocardiography or other advanced modalities to determine whether or not this played a role with regard to the outcomes.

Third, the sublingual microvascular flow index, which was selected as a microcirculatory endpoint for this study, does not necessarily match with the microvascular perfusion in the gastric conduit or other vital organs. Past studies have indicated that there exists a possibility of divergence between microvascular perfusion in the intestine and that in the sublingual tissues in severe sepsis patients. Moreover, there was no measurement of microvascular perfusion in the gastric tube. The time course of the microvascular flow index was not monitored in this study, apart from the selected time points. It was also not correlated with postoperative outcomes.^9,10

Fourth, as far as the microvascular fluid responsiveness parameters are concerned, which were studied and evaluated in this research, these parameters were derived from an analogue model based on arterial waveform and related data. Although past studies have indicated a fair correlation and tracking of microvascular fluid responsiveness parameters, there exists a possibility of coupling or systematic bias.¹¹

Pregnant patients and their procedures, as well as those involving pregnant patients, were excluded from this study. As such, it should be noted that the study findings concerning pregnant patients are only extrapolated. Although the major finding concerning venous congestion and microvascular perfusion, as well as macrovascular adequacy, appears physiologically plausible, this finding would most probably be applicable to a broad range of patients.

Future Directions

Building on this trial, several lines of investigation and quality‑improvement (QI) work would be highly relevant for anesthesia trainees, including those interested in obstetric practice:

Prospective intraoperative study of CVP and microcirculatory surrogates in high‑risk obstetric patients.

The trainees could design a small study with patients suffering from severe preeclampsia or cardiomyopathy and requiring a cesarean section or a non-obstetric surgical procedure. The study would correlate CVP and MAP with readily measurable indices of microcirculatory function, such as lactate curves or skin/mucosal perfusion indices, or even near-infrared spectroscopy if this technology is available. The purpose would be to ascertain if this disconnection between macro- and microcirculatory function, which we have seen with esophagectomy patients, also occurs in obstetric patients.

QI project to implement CVP “upper limit” alerts during high‑risk abdominal or pelvic surgery.

Following on from the knowledge that CVP  ≥  23 mmHg is associated with poor microcirculatory flow, a quality improvement project could be undertaken by a group of trainees to develop local guidelines to address this. This would mean that, should CVP rise to a certain level, it is not simply a matter of “all is well,” but a more urgent intervention is necessary. This could be done by reviewing fluid status, ventilator settings, and vasopressor use, with subsequent review to assess feasibility and acceptance.

Educational and protocol‑development initiative in obstetric anesthesia.

The trainee can play a role in the production of an educational module and protocol for labor and delivery operating rooms, which centers upon microcirculation-conscious hemodynamics. This is achieved by judicious volume administration, tailored vasopressor therapy, and accounting for conditions that increase venous pressure, such as positive pressure ventilation, pneumoperitoneum in non-obstetric laparoscopic procedures, and Trendelenburg position. Esophagectomy cases can be used as a mechanistic case study to discuss the rationale behind the inadequacy of a “normal” mean arterial pressure to guarantee the safety of vulnerable microcirculatory beds.

Multicenter observational registry combining macro‑ and microcirculatory data.

Such collaboration could also help in the creation of a registry of high-risk surgical patients with standard data on central venous pressure (CVP), mean arterial pressure (MAP), cardiac output (CO), and, if available, data on microcirculatory flow in the sublingual tissues or other accessible tissues. Such a registry would be instrumental in validating whether the relationships between CVP, mean systemic filling pressure (MSFP), and MFI observed in this study are present in a broader and more diverse group of patients and are related to important outcomes such as leakage and organ dysfunction.

These activities are appropriate for medical students and residents, align with VJM's mission to advance health and equity, and directly build on the mechanistic findings from this trial regarding venous congestion and microcirculatory flow.

---

References

1. He H, Gruartmoner G, Ince Y, et al. Effect of pneumoperitoneum and steep reverse-Trendelenburg position on mean systemic filling pressure, venous return, and microcirculation during esophagectomy. J Thorac Dis. 2018;10(6):3399-3408. doi:10.21037/jtd.2018.05.169

2. Vellinga NA, Ince C, Boerma EC. Elevated central venous pressure is associated with impairment of microcirculatory blood flow in sepsis: a hypothesis generating post hoc analysis. BMC Anesthesiol. 2013;13:17. doi:10.1186/1471-2253-13-17

3. Chen CY, Zhou Y, Wang P, Qi EY, Gu WJ. Elevated central venous pressure is associated with increased mortality and acute kidney injury in critically ill patients: a meta-analysis. Crit Care. 2020;24(1):80. doi:10.1186/s13054-020-2770-5

4. Li DK, Wang XT, Liu DW. Association between elevated central venous pressure and outcomes in critically ill patients. Ann Intensive Care. 2017;7(1):83. doi:10.1186/s13613-017-0306-1

5. Eskesen TG, Wetterslev M, Perner A. Systematic review including re-analyses of 1148 individual data sets of central venous pressure as a predictor of fluid responsiveness. Intensive Care Med. 2016;42(3):324-332. doi:10.1007/s00134-015-4168-4

6. Magder S. Understanding central venous pressure: not a preload index? Curr Opin Crit Care. 2015;21(5):369-375. doi:10.1097/MCC.0000000000000238

7. Gelman S. Venous function and central venous pressure: a physiologic story. Anesthesiology. 2008;108(4):735-748. doi:10.1097/ALN.0b013e3181672607

8. Yartsev A. Circulatory changes in pregnancy | Deranged Physiology. Accessed March 18, 2026. https://derangedphysiology.com/main/required-reading/obstetric-intensive-care/Chapter-313/circulatory-changes-pregnancy

9. Boerma EC, van der Voort PHJ, Spronk PE, Ince C. Relationship between sublingual and intestinal microcirculatory perfusion in patients with abdominal sepsis. Crit Care Med. 2007;35(4):1055-1060. doi:10.1097/01.CCM.0000259527.89927.F9

10. Edul VSK, Ince C, Navarro N, et al. Dissociation between sublingual and gut microcirculation in the response to a fluid challenge in postoperative patients with abdominal sepsis. Ann Intensive Care. 2014;4:39. doi:10.1186/s13613-014-0039-3

11.        De Santis P, De Fazio C, Franchi F, et al. Incoherence between Systemic Hemodynamic and Microcirculatory Response to Fluid Challenge in Critically Ill Patients. J Clin Med. 2021;10(3):507. doi:10.3390/jcm10030507