Arch Gynecol Obstet (2014) 290:1041–1044 DOI 10.1007/s00404-014-3355-9
Cerebral oedema following robotic surgery: a rare complication Chloe Barr • Thumuluru Kavitha Madhuri Pradeep Prabhu • Simon Butler-Manuel • Anil Tailor
Received: 17 March 2014 / Accepted: 1 July 2014 / Published online: 6 August 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Introduction We present an unusual complication following robotic assisted radical hysterectomy. Case report A 51-year-old female with stage 1B1 cervical cancer underwent a robotic assisted radical hysterectomy. The procedure was prolonged with difficulties dissecting the left parametrium and vaginal fornix with persistent bleeding from the left vaginal vault. Post-operatively the patient was electively sedated and ventilated. Extubation was difficult due to patient agitation but achieved on day 2. Agitation persisted and a head CT scan was performed and a diagnosis of cerebral oedema was made. Discussion Factors contributing to this case include prolonged operating time, prolonged Trendelenburg position with high pressures of CO2 pneumoperitoneum and excessive blood loss. These factors may contribute to poor cerebral venous outflow, increasing intracranial pressure leading to increased risk of cerebral oedema. Conclusion The mechanics of robotic assistance may be used to reduce these risks by significantly reducing intraabdominal pressure improving venous return. Summary The use of robotics in surgery has been increasing over the last 10 years, and the benefits have been well documented. We present an unusual complication following robotic assisted radical hysterectomy performed for cervical cancer.
C. Barr T. K. Madhuri (&) S. Butler-Manuel A. Tailor Department of Gynaecological Oncology, Royal Surrey County Hospital NHS Foundation Trust, Egerton Rd, Guildford GU2 7XX, UK e-mail: [email protected]
P. Prabhu Department of Anaesthetics and Intensive Care, Royal Surrey County Hospital NHS Foundation Trust, Egerton Rd, Guildford GU2 7XX, UK
Introduction The use of robotics in gynaecological surgery was approved by the US drug and food administration (FDA) in 2005 and its use in gynaecological surgery has been increasing ever since . Robotic surgery requires the patient to be positioned in steep Trendelenburg, with a CO2 pneumoperitoneum (Fig. 1), the combination of which presents several anaesthetic challenges and contributes to its limitations. We present an unusual complication following robotic assisted radical hysterectomy performed for cervical cancer.
Case report A 51-year-old multiparous lady, diagnosed with stage 1B1 adenocarcinoma of the cervix was listed for a robotic assisted radical hysterectomy (RRH), bilateral salpingo-oopherectomy and bilateral pelvic lymph node dissection. She had a complex medical history including chronic lymphocytic leukaemia (CLL), which was diagnosed 9 years ago, and paranoid schizophrenia. Preoperative assessment revealed a weight of 98 kg, a height of 162 cm, giving her a BMI of 37. Her heart rate was 90 beats per minute and she had normal oxygen saturations. Her routine blood tests were all within normal parameters, and ECG showed she was in sinus rhythm. Her blood pressure was raised at 155/90 mmHg. She was not on any anti-hypertensive medication at the time, only analgesics and risperidone injections. In the absence of LVH on ECG, her hypertension was attributed to white coat syndrome. The patient was induced using intravenous Propofol 200 mg, Fentanyl 100 mg and Rocuronium 50 mg.
Fig. 1 A patient positioned in steep Trendelenburg
Maintenance of anaesthesia was achieved using air, O2 Sevofluorane and a Remifentanil infusion. No regional block or epidural was placed. She was intubated orally using a size seven endotracheal tube. In addition to mandatory monitoring, she had a central line placed in the right internal jugular vein for central venous pressure monitoring and an arterial line placed in her radial artery for continuous arterial blood pressure monitoring. The patient was placed in lithotomy position. The uterus was not manipulated, and a rectal bougie was used in the vagina to delineate the fornices. She was then positioned in 30° Trendelenburg and pneumoperitoneum was achieved using a Veress needle, to a pressure of 16 mmHg. A Da Vinci S robot (Intuitive Surgical Inc) was docked on the patient’s right side using four arms. Bilateral pelvic lymphadenectomy was the first step in the procedure and was uncomplicated, lasting 60 min. The radical hysterectomy proceeded slowly due to difficulties dissecting the left parametrium and left vaginal fornix. Difficulties were again encountered during closure of the vaginal vault due to persistent bleeding from the left vaginal angle. Six hours into the procedure, the anaesthetic team noticed the patient had started to develop facial oedema. The anaesthetic monitoring showed high peak airway pressures, a high endtidal (ET) CO2 and progressive hypoxia despite increasing oxygen supplementation. To achieve haemostasis and in view of the deteriorating anaesthetic parameters, the decision was taken to perform a Pfannenstiel incision to achieve haemostasis and complete the surgery which was achieved uneventfully. The total amount of time the patient spent in steep Trendelenburg position was 7 h. The total blood loss was 1,600 ml, and the patient was fluid resuscitated intra-operatively with 1,500 ml VolplexÒ and 4,000 ml Ringer-Lactate solution for injection (Hartmann’s solution). Post-operatively she also received a transfusion of two units of packed cells in the intensive care unit (ICU) over the next 48 h.
Arch Gynecol Obstet (2014) 290:1041–1044
The patient was sedated and electively ventilated overnight. Extubation was attempted on day 1 but was abandoned as she became very agitated. On day 2 the patient was successfully extubated, although she remained agitated. The patient also developed a tachycardia and pyrexia. A CT scan of the abdomen and pelvis was arranged which excluded any focus of intra-abdominal sepsis. In view of the facial oedema and continuing agitation, a CT head scan was also performed on day 2, which revealed normal ventricles and other cerebrospinal fluid spaces with no acute haemorrhage or extra-axial collections. A lumbar puncture was normal. A presumptive diagnosis of cerebral oedema was made, and the patient was managed conservatively with supportive care on ICU where she remained for 5 days. She was then transferred to the surgical ward for continued observation and mobilisation. She was discharged on day 11 post-op in good health. She has not developed any long-term sequelae and has no sign of recurrence of her cervical cancer after more than 4 years of clinical follow-up.
Discussion Robotic surgery offers many potential advantages, both for the patient and the surgeon. Advantages for the patient include reduced blood loss intra-operatively, reduced postoperative pain, shorter hospital stay and faster post-operative recovery . For the surgeon, the robot provides increased surgical precision due to 3-D high definition view of the surgical field and wristed instruments, which offer a huge range of movement . Many of the limitations and complications of robotic surgery are associated with the positioning of the patient in steep Trendelenburg while maintaining a pneumoperitoneum, and/or with prolonged operative procedures. The most frequently reported complications include: brachial plexus injury , myocardial infarction , facial oedema, laryngeal oedema causing post-operative respiratory distress , profound ischaemia of the lower limbs  and intestinal and urological injuries . Cerebral oedema is an excess accumulation of fluid in the intra or extracellular spaces in the brain and may be classified according to cause (Table 1) . This is a very uncommon complication following robotic surgery, with only two previously documented cases . The exact mechanisms of why cerebral oedema may develop following this type of surgery are unknown, and studies have shown that the cerebral physiology remains within safe limits of autoregulation in such extreme conditions . Here we discuss some of the contributing factors in Table 1. This case was one of the first RRH performed in our unit, and so occurred early in the learning curve of robotic
Arch Gynecol Obstet (2014) 290:1041–1044 Table 1 Classification of cerebral oedema Classification Vasogenic
Disruption of the cerebral capillary is the underlying mechanism, which allows fluid to leak into the extravascular space
Derangement in cellular metabolism leads to inadequate functioning of the sodium/potassium pump. As a result there is cellular retention of sodium and water
Extra cellular fluid osmolality in the brain is usually greater than that of plasma. When the plasma is diluted there is movement of water into the brain down the abnormal gradient
Results from direct pressure on the capillaries with transudation of fluid into the extracellular fluid
surgery at our institution. The relative unfamiliarity of a new approach for the procedure, combined with her obesity, and the fact that there were two complex procedures being performed, caused the operation to take far longer than is usual. The Trendelenburg position causes an elevation in venous pressure in the brain due to the gravitational effect on the venous drainage. This leads to increased cerebral blood volume and increased cerebrospinal fluid volume (CSF). This in turn leads to increased intracranial pressure (ICP). Following a study of 15 patients, Mavrocordatos et al. reported that ICP can increase from 8.8 to 13.3 mmHg with 30° of Trendelenburg . It is likely that the cerebral oedema develops due to movement of extracellular fluid from the vessels into the surrounding brain parenchyma. During laparoscopy and robotic procedures, the abdomen is insufflated using CO2 to create a potential space within which the surgery may be performed. We used a pneumoperitoneum of 16 mmHg in this case: a significant rise over normal intra-abdominal pressure. This may obstruct venous return from the lumbar venous plexus and can contribute to a raised ICP . The CO2 used to inflate the abdomen can dissolve into the bloodstream causing hypercarbia . High levels of arterial CO2 may cause cerebral vasodilation leading to increased cerebral blood flow and raised ICP. Hypoxia may also contribute to a raised ICP in this way. In our case, hypoxia was not an immediate issue as oxygenation was well controlled and regulated. However, peri-operative end-tidal CO2 monitoring demonstrated that the patient did have increasing levels of CO2 (Fig. 1). Rosenthal et al. have shown that a combination of increased intra-abdominal pressures of 16 mmHg and Trendelenburg position increase the ICP 150 % over the control level in pigs . It is well known that the volume expanding effects of crystalloid are short lived as they move quickly from the intravascular to the extravascular compartment. Excessive
crystalloid resuscitation, therefore, may lead to oedema. The dependent effects of operating in Trendelenburg position mean that the oedema has a tendency to develop around the face, eyes and larynx making extubation difficult in some cases and contributing to cerebral oedema . Recent studies have recommended that patients should be fluid restricted during robotic procedures to try to prevent the associated upper airway and facial oedema [2, 12, 13]. The case described occurred early on in our learning curve—this was the first robotic radical hysterectomy performed by this surgeon 4 years ago—leading to the patient being in Trendelenburg position for 7 h. The use of robotics is increasing worldwide, and will play a big part in the future of minimal access surgery. As with any new technology or surgical technique, there is a learning curve, and this is well documented . This will inevitably lead to increased operating times early on in the surgeon’s experience. What is unclear from the literature is the maximum duration we should be allowing patients to be in the Trendelenburg position. As our experience using robotics has progressed, the surgical procedure time has inevitably decreased. Early on in the learning curve, however, converting to an open procedure should be considered at a certain point in time to avoid complications as occurred with this case. Schramm et al. looked at cerebrovascular autoregulation during the steep Trendelenburg and pneumoperitoneum required a robotic assisted prostatic surgery and found that with longer periods of Trendelenburg there is deterioration in cerebrovascular autoregulation, leading to increased risk of cerebral hyperaemia and cerebral oedema . There is no absolute time line or duration of procedure described in the literature, perhaps because there are many physiological variables to consider, as described previously, including the complexity of the surgery. In the meantime, changes can be made in an attempt to reduce the risk of this complication. First and foremost, there should be a strict restriction on fluids, particularly crystalloid fluids. This is a variable that can be easily controlled and monitored and should be done in all cases. The pressure of the pneumoperitoneum can be reduced from 16 to only 8 mmHg. This surprisingly low pressure pneumoperitoneum is made possible due to the mechanical assistance offered by the robot which supports the weight of the abdominal wall. The degree of Trendelenburg can be reduced, and we currently use a Trendelenburg of strictly 30°. A Trendelenburg of 30° combined with a lower pressure of pneumoperitoneum allows good surgical views without anaesthetic compromise has been demonstrated in the literature . During the learning phase, an attempt should be made to try and avoid prolonged procedures. There should be an awareness of the length of time of the ongoing procedure and, perhaps one could consider reviewing the case every 3 h
and either undock the robot, allowing normalisation of the position and a reduction of the pneumoperitoneum enabling correction of the physiological parameters before continuing, or abandoning the procedure at this point. Support and guidance from an appropriately trained surgical proctor may reduce the console time in the early learning phase of such complex procedures, and this was not available to us at the time.
Conclusion This rare case of cerebral oedema following robotic surgery was associated with prolonged surgery in the early learning curve of this modality, a steep Trendelenburg position, high intra-abdominal CO2 pressure and fluid resuscitation. Fluid restriction and the mechanics of robotic assistance may be used to reduce these risks by significantly reducing intra-abdominal pressure improving venous return. There should be an awareness of this complication in complex surgical procedures, particularly early in the learning curve.
Conflict of interest All the authors confirm and declare that they have no conflict of interest.
References 1. Madhuri TK, Hamzawala I, Tailor A, Butler-Manuel S (2012) Robot assisted surgery in gynaecologic oncology––starting a program and initial learning curve from a UK tertiary referral centre: the Guildford perspective. Int J Med Robot. 8(4):496–503 2. Phong S, Koh L (2007) Anaesthesis for robotic assisted radical prostatectomy: considerations for laparoscopy in trendelenburg position. Anaesth Intensiv Care 35(2):281–285
Arch Gynecol Obstet (2014) 290:1041–1044 3. Thompson J (2009) Myocardial infarction and subsequent death in a patient undergoing robotic prostatectomy. AANA J. 77(5):365–371 4. Horgan A, Geddes S, Finlay I (1999) Lloyd-davies position with trendelenburg––a disaster waiting to happen? Dis Colon Rectum 42:916–919. doi:10.1007/BF02237102 5. Pereira Arias JG, Gamarra Quintanilla M, Leibar Tamayo A, Astobieta Odriozola A, Ibarluzea Gonza´lez G (2010) Complications and incidences in our first 250 robotic radical prostatectomies. Actas Urol Esp 34(5):428–439 6. Marmarou A (2004) The pathophysiology of brain edema and elevated intracranial pressure. Clevel J Med 71(S1):S6–S8 7. Pandey R, Garg R, Darlong V, Punj J, Kumar A (2010) Unpredicted neurological complications after robotic laparoscopic radical cystectomy and ileal conduit formation in steep trendelenburg: two case reports. Acta Anaesth Belg 61:163–166 8. Kalmar A, Foubert L, Hendrickx J, Mottrie A, Absalom A, Mortier E, Struys M (2010) Influence of steep trendelenburg position and CO2 pneumoperitoneum on cardiovascular, cerebrovascular and respiratory homeostasis during robotic prostatectomy. BJA 104(4):433–439 9. Mavrocordatos P, Bissonnette B, Ravussin P (2000) Effects of neck position and head elevation on intracranial pressure in anaesthetised neurosurgical patients. J Neurosurg Anaesthesiol 12(1):10–14 10. Halverso A, Buchanan R, Jacobs L, Shayani V, Hunt T, Riedal C, Sackier J (1998) Evaluation of mechanisms of increased intracranial pressure with insufflation. Surg Endo SC 12:266–269 11. Gupta K, Mehta Y, Sarin Jolly A, Khanna S (2012) Anaesthesia for robotic gynaecological surgery. Anaesth Intensive Care 40:614–621 12. Gainsburg D (2012) Anaesthetic concerns for robotic assisted laparoscopic radical prostatectomy. Minerva Anestesiol 78(5):596–604 13. Sullivan M, Frost E, Lew M (2008) Anaesthetic care of the patient for robotic surgery. MEJ Anesth 19(5):967–982 14. Schramm P, Treiber AH, Berres M et al (2014) Time course of cerebrovascular autoregulation during extreme trendelenburg position for robotic assisted prostatic surgery. Anaesthesia 69(1):58–63 15. Madhuri TK, Butler-Manuel SA (2011) Robotic surgery in gynaecology. ONJA 6(3):96–97