Short summary

Perioperative haemodynamic optimization can improve outcome in high-risk surgery but the optimum intravenous fluid to use has yet to be determined. In the current study 60 patients undergoing elective major abdominal cancer surgery are randomly assigned to receive either colloid (5 % Human Albumin) or balanced crystalloid (Ringerfundin) as haemodynamic optimization fluid. Changes in intraoperative global and local (mesenteric) oxygen delivery, other haemodynamic variables as well as postoperative outcome variables are compared.

---

Rationale

Evidence suggests that goal directed therapy, also known as haemodynamic optimization, improves outcome when used during the perioperative period in high-risk abdominal surgery. Cardiac output and other haemodynamic parameters are optimized to predefined goals by administration of intravenous fluids and vasopressors or inotropic agents, aiming to improve tissue perfusion and finally oxygen delivery, and thereby facilitate tissue-recovery.

The current evidence is based on studies of synthetic colloid administered as bolus fluid in a haemodynamic optimization protocol. In this setting colloid is believed to have a theoretical advantage to crystalloid, since lower volumes are needed to achieve similar haemodynamic endpoints. Studies have indicated that crystalloid can be damaging to the endothelial glycocalyx resulting in increased vascular permeability, thus leading to a predisposition to tissue oedema. It is well documented, that excess volumes of intravenous fluid in the perioperative period during colorectal surgery is associated with a higher rate of complications, while fluid-restriction do not seem to be beneficial either.

In contrast, a higher risk of acute renal failure and an increased mortality is observed among patients with sepsis receiving synthetic colloid (HES 6 %) in the intensive care unit. Thus, it has become clinical practice in many institutions to administer balanced crystalloid as the sole agent for perioperative intravenous fluid-therapy based on studies of fluid resuscitation in critical illness. Still, it is uncertain, whether results based on a study of patients with critical illness can be extrapolated to patients undergoing elective surgery. At the same time it is believed, but not well documented, that albumin, as a human plasma derived product, is without the same adverse effects as described with synthetic colloid. However, a recent observational study have described, that both synthetic colloid (HES 6 %) and albumin (Human Albumin 5 %) is associated with an increased risk of acute renal failure in an elective perioperative orthopaedic setting.

Only a few prospective trials have compared colloid and crystalloid in the perioperative setting, and no clear conclusions about the usefulness or safety of either colloids or crystalloids can be drawn.

In summary, the optimum type of intravenous fluid to use when performing perioperative haemodynamic optimization remains to be determined. The current study aims to help answer this question.

If the use of colloid as a haemodynamic optimizing fluid is beneficial to the overall outcome of a high-risk surgery population, then it would be a result of colloids superior effect on perioperative haemodynamic stability and oxygen delivery. Consequently, the first step is to clarify the effect of colloid and crystalloid on perioperative systemic and regional oxygen delivery as well as haemodynamic stability.

Our study is a prospective randomized controlled trial comparing the effect of albumin (Human Albumin 5 %) with balanced crystalloid (Ringerfundin) when used in a perioperative goal-directed haemodynamic algorithm during major abdominal cancer surgery hypothesizing albumin being superior to balanced crystalloid in achieving the described outcomes.

Main objective:

To investigate differences in intraoperative global and local (mesenteric) oxygen delivery.

Secondary objectives:

To investigate differences in intraoperative hemodynamic stability (cardiac output, stroke volume, mean arterial pressure, systemic vascular resistance, blood gas values, need of inotropic support), fluid balance, body weight, fluid related complications and length of stay in the ICU and the hospital for 30 days postoperatively.

---

Description of the cohort

Inclusion criteria:

• Patients eligible for relevant surgery; oesophagectomi, total gastrectomy, The Whipple procedure (pancreatoduodenectomy), total pancreatectomy.
• Age over 18 and under 80 years.

Exclusion criteria:

• Preoperative renal failure (eGFR < 30 or renal replacement therapy)
• Lithium treatment
• Body weight < 40 kg
• Not sinus rhythm
• Aortic valve regurgitation
• Known allergic reactions to albumin
• Patients with pancreatic cancer, who have been down-staged using chemotherapy and/or radiation therapy
• Patients with unresectable disease
• Pregnancy
• Patients declining informed consent

Data and biological material

Baseline patient characteristics:

• Age, gender, body height and weight.
• ASA-score, POSSUM-score
• Chemotherapy before surgery
• Type and duration of surgery
• Blood tests; haemoglobin, platelets, albumin, creatinine, coagulation factor, liver enzymes, bilirubin.

Intraoperative variables:

• Cardiac Index (l/min/m²)
• Stroke Volume Index  (ml/min/m²)
• Systemic Vascular Resistance Index (dynes * s/cm⁵/m²)
• Pulse Pressure Variation (%)
• Oxygen Delivery Index (ml/min/m²)
• Mean Arteriel Pressure (mmHg)
• Heart Rate (beats/min)
• Central Venous Pressure (mmHg)
• Dosage of vasopressor and inotrope agents (ug/kg/min)
• Arterial blood gases; base excess, lactate, haemoglobin (mmol/l)

Intraoperative ultrasound measurements of the superior mesenteric artery:

• Arterial diameter (mm²)
• Time average mean, TAM (cm/s)
• Time average maximum, TAMX (cm/s)
• Volume flow (ml/min)
• Mesenteric oxygen delivery (ml/min)

Intraoperative and postoperative fluid balance:

• Blood loss, gastric aspirate, perspiration, urine output, drainage fluids (ml)
• Administration of intravenous crystalloids and colloids, blood products, enteral nutrition and fluids (ml)
• Body weight

Other postoperative variables:

• Gastrointestinal function; time to first flatus and stool (days)
• Length of stay in the ICU (ours) and in the hospital (days)

Complications during hospital admission:

• Pneumonia
• Wound infection
• Acute myocardial ischemia
• Acute kidney injury
• Atrial fibrillation
• Pulmonary oedema
• Intra-abdominal hypertension
• Anastomotic leakage
• Need of inotropic support
• Need of mechanical ventilation
• Need of acute dialysis

Collaborating researchers and departments

Department of Anaesthesia and Intensive Care Medicine, Odense University Hospital

• Investigator Jannie Bisgaard Stæhr, MD, PhD
• Investigator Rasmus Madsen, MD
• Sponsor Anders Gadegaard Jensen, MD, DMSc

Department of Surgery, Odense University Hospital

• Surgical adviser Michael Bau Mortensen MD, DMSc