The use of orthotopic liver transplantation (OLT) has considerably changed over just a few years and it has now become the only definitive treatment for acute and chronic end-stage liver diseases. Over the last 10 years, OLT-related mortality and complication rates have significantly decreased and the mean 1-year survival rate is now more than 90% as a result of improved operating techniques, the consolidation of basic knowledge and experience, better patient selection and preparation, and the use of innovative drugs and technologies .
Furthermore, the perioperative management of OLT patients has also evolved: for example, until very recently, a postoperative period of mechanical artificial ventilation was considered a standard of care for anyone undergoing OLT because it was believed that reducing the stress associated with arousal from anaesthesia and the resumption of spontaneous ventilation contributed to ensuring a soft and controlled transition from the surgical to the postoperative recovery phase . However, this approach has been questioned because, especially when used with positive end-expiratory pressures, mechanical ventilation affects blood flow in the splanchnic region by increasing venous pressure, and the reduced venous outflow causes congestion of the graft, which may thus be negatively affected precisely at the crucial time in which it is recovering from the inevitable ischaemia-reperfusion damage due to its hypothermal conservation [3,4]. In addition to its effects on specifically physiopathological factors, reducing or abolishing post-OLT mechanical ventilation may offer economic and organizational advantages by shortening the patients' length of stay in an intensive care unit (ICU) .
We have previously shown that early tracheal extubation (i.e. within 3 h of the end of surgery) can be safely performed in a large number of unselected OLT patients . We here describe the results of a study designed to evaluate the possibility of safe and reliable extubation in the operating theatre immediately after surgery.
All of the subjects undergoing OLT at our centre between 1 June 1999 and 31 May 2004 were admitted to the study (which was approved by our Institutional Review Board) after having given their informed consent. The exclusion criteria were OLTs due to acute hepatic failure, emergency re-transplantations, death during the procedure or the first 24 postoperative hours, and the need for mechanical ventilation before transplantation.
Anaesthesia and surgical management
The anaesthesia technique was the same in all cases: induction with intravenous (i.v.) fentanyl 0.2 mg, sodium thiopental and cisatracurium, and maintenance with sevoflorane in a 50% air/oxygen low-flow respiratory mixture, remifentanil (0.2-0.3 μg kg−1 min−1) and cisatracurium (3 μg kg−1 min−1). Haemodynamic monitoring included invasive systemic arterial pressure and the use of a pulmonary artery catheter (CCO/SVO2 Thermodilution Catheter, Edwards Life Sciences LLC, Irvine, CA, USA). An extracorporeal veno-venous bypass between the portal and inferior vena cava and the superior vena cava was routinely used in all of the procedures.
The immunosuppressive protocol included oral cyclosporin A (CsA: Sandimmun Neoral®; Novartis Pharma S.A., Huningue, France) at a dose of 15 mg day−1 on the day of surgery, subsequently titrated to maintain blood trough levels of 200-250 ng dL−1; methylprednisolone (Solu-Medrol®; Pharmacia & Upjohn, Puurs, Belgium) at an intraoperative dose of 10 mg kg−1, subsequently reduced by 50% day−1 to a prednisolone dose of 20 mg day−1; and basiliximab (Simulect®; Novartis Pharma S.A., Huningue, France) at an i.v. dose of 20 mg during the anhepatic phase and on the 4th postoperative day.
Perioperative anti-infective prophylaxis was based on the administration of cefazolin (1 g 3 times a day) and ampicillin-sulbactam (2 + 0.2 g twice a day). Postoperative pain was controlled by administering intra muscular (i.m.) morphine 0.1 mg kg−1 40 min before the end of the procedure, followed by a continuous i.v. infusion of 20-40 mg day−1 from the time of the patient's arrival in the ICU. Humidified oxygen (3-5 L min−1) was given to the extubated patients through a nasopharyngeal tube and a respiratory physiotherapy programme was started as soon as possible. In the case of mechanical ventilation, positional manoeuvres were performed routinely. Complicated patients received antibiotics, and renal, haemodynamic and respiratory support as appropriate. Organ dysfunction was defined on the basis of generally accepted criteria .
Tracheal extubation protocol
The patients were extubated in the operating theatre at the end of surgery on the basis of standardized and universally accepted criteria: that is, patient awake (able to follow simple orders, spontaneous eye opening), clinical evidence of neuromuscular reversal (ability to lift the head and swallow; a tidal volume of >6 mL kg−1), normocarbia (ETCO2 30-40 mmHg), a respiratory rate of <25 min−1 and good oxygenation (SPO2 >95%, with FiO2 <0.5). Tracheal extubation was carried out under conditions of haemodynamic stability as judged by the attending anaesthetist and all of the patients were admitted to the ICU.
The collected data included patient characteristics data, the aetiology of the liver disease indicating OLT, the preoperative Child-Pugh score, the Model for End-stage Liver Disease (MELD) score and United Network for Organ Sharing (UNOS) score; surgical and anaesthesia management (cold ischaemia and operating times, transfusion requirements), postoperative ventilation times (for those not immediately extubated), length of ICU stay, the need for re-intubation during the 24 h after the first attempt at postoperative extubation and any postoperative complications.
The patient characteristics data and clinical variables characterizing the studied population are expressed as mean ± SD and percentage frequencies. The groups were compared using the χ2-test and one-way analysis of variance (ANOVA) with Bonferroni's correction for multiple comparisons. The MELD score predicting immediate extubation (the score with the best sensitivity/specificity ratio) was calculated by means of receiver operator characteristic area under the curve (ROC AUC) analysis. The statistical analysis was made using STATA software (College Station, TX, USA), with a 5% significance level (0.05).
During the study period, 364 patients underwent 377 OLTs: as 10 patients and 12 procedures were excluded on the basis of the protocol criteria, the analysis was based on 354 patients and 365 transplantations.
Two hundred and seven patients (58.5%, Group A) underwent 211 immediate tracheal extubations: the mean time between the end of surgery and extubation was 0.4 ± 1.4 min (range 0-12); in 110 patients (31%) undergoing 113 OLTs, postoperative ventilation lasted less than 24 h (Group B), including 77 (70% of the group) in whom it lasted for <3 h; the remaining 37 patients (10.5% undergoing 41 OLTs) required prolonged ventilatory assistance (Group C) (Table 1). The annual number of procedures after which immediate tracheal extubation was possible increased considerably during the study period and, in the last year, reached 82.5%; the number of procedures requiring intermediate ventilation (up to 24 h) simultaneously decreased (Table 2). Finally, the number of immediate extubations per individual member of the team of anaesthetists also increased over time: initially, only one of them extubated patients immediately after OLT but, at the end of the study period, the number of such procedures was equally distributed (Fig. 1).
The study groups were different in terms of age (Group C patients were younger), gender (Group A had a higher proportion of males) and the severity of the underlying liver disease as measured by the MELD score (Table 1). The results of the ROC analysis showed that a MELD score of 11 had the best sensitivity/specificity ratio as a threshold value for identifying the subjects with a greater possibility of being immediately extubated (Table 3). Figure 2 shows the plot of the ROC AUC analysis for a MELD score of 11 and its significant predictive power (ROC AUC = 0.61; 95% CI = 0.54-0.66; P < 0.05). The patients were homogeneous in terms of their Child-Pugh scores (the most severe category C patients were equally distributed in the three groups: P = 0.4); UNOS class 1 and 2b patients were equally distributed in Groups A and B (P = 0.2), but represented a higher proportion in Group C patients (P < 0.001).
In relation to intraoperative events, the groups were homogeneous in terms of the duration of cold ischaemia and surgery, but intraoperative transfusions were more frequent in Group C. The number of patients experiencing severe intraoperative haemorrhage (defined as a need for 10 or more units of concentrated red blood cells) was different in the three groups: 5 in Group A, 11 in Group B and 13 in Group C (P < 0.001), and respectively 6, 11 and 14 patients (P < 0.001) experienced a reduction in blood pressure (BP) requiring the administration of vasoactive drugs at the time of graft reperfusion (Table 1).
Six patients (1.7%) were re-intubated during the course of the study: two in Group A (one on postoperative day 2 because of the surgical revision of haemostasis and one on postoperative day 3 because of pneumonia due to Pseudomonas aeruginosa), and four in Group C (two because of haemostasis revision on postoperative days 2 and 3, one because of hepatic artery thrombosis on postoperative day 10, and one because of the onset of pneumonia due to Klebsiella pneumoniae on postoperative day 5). Seventeen subjects (4.8%) underwent non-invasive mechanical ventilation administered through a mask or helmet in accordance with previously published criteria and indications : 11 in Group A (for a mean duration of 69.8 ± 25 h: range 24-120) and six in Group B. None of these patients were re-intubated.
Seventeen (4.8%) of the 354 study subjects died during their first hospitalization: two in Group A (one massive pulmonary embolism, and one myocardial infarction), four in Group B (one due to the sequelae of a cerebral thrombosis, one because of septic shock and two cases of multi-organ dysfunction) and 11 in Group C (five cases of multi-organ dysfunction, three of septic shock, two of primary non-function of the transplanted organ and one due to the sequelae of cardiocirculatory arrest occurring during the phase of graft reperfusion) (P < 0.001 vs. Groups A and B). The other complications recorded during the ICU stay are shown in Table 4.
It was commonly believed for many years that, together with adequate sedation, prolonging mechanical ventilation beyond the surgical phase of a complex procedure effectively contributes to limiting postoperative stress [9,10]. However, mechanical ventilation does not always benefit all post-surgical patients because the lower intrapleural pressure during spontaneous ventilation facilitates venous return and favours end-diastolic filling of the heart, cardiac output and hepatic flow [11,12], and this can improve venous drainage and liver circulation [13,14] especially when it is useful to create the best haemodynamic conditions for functional graft recovery. Consequently, the common practice of ensuring that all OLT patients undergo a period of postoperative mechanical ventilation has recently been questioned by various authors describing albeit smaller case series than ours [5, 15-17].
During the course of our study, we found it possible to undertake immediate tracheal extubation after the end of surgery (0.4 ± 1.4 min after completing the medication of the surgical wounds) in 58.5% of cases; however, analysis of the data by each study-year showed that this percentage increased substantially from 18.9% in the first year to 44.8% in the second and to as much as 82.5% in the last year. This progression may not only mirror the learning curve of the team of anaesthetists, but also a curve of ‘confidence’ in extubating patients in the operating theatre immediately after OLT. A spirit of emulation seems to have been another important factor insofar as 80% of the operating theatre extubations in the first year were performed by only one anaesthetist whereas, by the end of the study, the number of immediate extubations was equally distributed among all of the team members. Over the same period of time, the number of subjects ventilated for a few hours after the procedure (Group B) decreased from 69.7% in the first year to 14.5% in the last year.
These data suggest that, before the beginning of the ‘era’ of immediate extubation, the prolongation of mechanical ventilation at our centre was probably mainly dictated more by reasons of convenience for the staff of the operating theatre (freed immediately after the operation) and ‘excessive’ caution or, rather, greater security for the team of anaesthetists, than by real clinical needs.
One fundamentally important aspect is the safety of extubation in the operating theatre. In this regard, although perhaps rather crude as a unit of measurement, the incidence of tracheal re-intubations within 24 h has the considerable advantages of clarity, simplicity and immediacy. In our experience, the total incidence of re-intubations in patients extubated immediately after OLT was 0.9% (only two patients). Eleven subjects in the same group (5.2%) received non-invasive ventilation because of some degree of postoperative respiratory failure: as previously underlined by Antonelli and colleagues , we found that this was an invaluable means of controlling such a dangerous event as hypoxaemia because none of them had to be re-intubated. Non-invasive ventilation is certainly well tolerated and, when administered in a timely manner, can efficaciously support a temporary deficit in respiratory function and thus significantly reduce the need for re-intubation.
Another central question is the possibility of predicting which patients can be extubated immediately and therefore also which can be potentially transferred to a ward or high dependency unit rather than the ICU. The severity of the pre-transplant hepatic disease may be a useful criterion , although we did not find the Child-Pugh score very reliable in this regard because there was no difference in the distribution of the most severely affected patients (Class C) in the three groups. On the contrary, the MELD score may provide some indications insofar as the patients with a score of more than 11 were more likely to require prolonged postoperative ventilatory support, just like those needing abundant transfusions during the course of surgery. However, although associated with a significant ROC AUC analysis, the sensitivity and specificity of this score are both only about 60%, and so even patients with higher scores may not necessarily need to be excluded from an immediate extubation protocol before the beginning of surgery.
In addition to its clinical benefits, the total abolition of post-OLT mechanical ventilation (but not just a reduction in its duration) [6,18] also allows economic savings due to the optimization of hospital resources by means of the re-allocation of ICU beds allowed by the shortening or elimination of ICU stays . In our experience, the use of immediate extubation of such a large number of OLT patients did not lead to any real reduction in the length of ICU stay. This is due to the organizational model of our centre, which lays down that all OLT patients be admitted to the ICU, which has a number of beds exclusively dedicated to liver transplant activities. The data and experience acquired during this study may lead us to review our protocols in order to begin examining the possibility of admitting immediately extubated OLT patients to our high dependency unit or directly to the surgical ward.
Finally, as immediate extubation is possible only if the patient is in very stable condition at the end of surgery, some preoperative (recipient selection) and intraoperative factors (an excellent and quick surgical technique, the correct management of blood coagulation, the maintenance of good thermal and metabolic homeostasis and the use of short-acting anaesthetic drugs) play a fundamental role. Successful immediate extubation could therefore be considered a significant indicator of the quality of intraoperative care and one of the main principles governing the management of anaesthesia for OLT .
In conclusion, our experience shows that a large proportion of liver transplant patients do not require any postoperative mechanical ventilation, the absence of which does not increase the risk of them developing postoperative complications attributable to immediate tracheal extubation. Furthermore, it is clear that evidence-based protocols in which resources are allocated exclusively on the basis of the real needs of patients can also be applied in the case of potentially highly critical and complex patients such as those undergoing OLT, regardless of the type of underlying disease. Given the increasing number of reports agreeing on these aspects, anyone involved in anaesthesia for liver transplantation should consider the emerging evidence also in the light of using immediate extubation as an important indicator of the quality of care.
1. Consensus Conference on indications of liver transplantation
2. McDonald M, Perkins JD, Ralph D, Carithers RL. Postoperative care. In: Maddrey WC, Sorrel MF, eds. Transplantation of the Liver
. Norwalk, USA: Appleton & Lange, 1995: 171-206.
3. Jullien T, Valtier B, Hongnat JM et al.
Incidence of tricuspid regurgitation and vena caval backward flow in mechanically ventilated patients. A color Doppler and contrast echocardiographic study. Chest
4. Carton EG, Plevak DJ, Kranner PW et al.
Perioperative care of the liver transplant patient: Part 2. Anesth Analg
5. Mandell MS, Lezotte D, Kam I, Zamudio S. Reduced use of intensive care after liver transplantation
: influence of early extubation. Liver Tranpl
6. Biancofiore G, Romanelli A, Bindi ML et al.
Very early tracheal extubation without predetermined criteria in a liver transplant recipient population. Liver Transpl
7. Deitch EA. Multiple organ failure; pathophysiology and potential future therapy. Ann Surg
8. Antonelli M, Conti G, Bufi M et al.
Noninvasive ventilation for treatment of acute respiratory failure in patients undergoing solid organ transplantation: a randomized trial. JAMA
9. Miller KA, Harkin CP, Bailey PL. Postoperative tracheal extubation. Anesth Analg
10. Siciliano D. Con: Early extubation is not preferable to late extubation in patients undergoing coronary artery surgery. J Cardiothor Vasc Anesth
11. Higgins JE. Pro: Early extubation is preferable to late extubation in patients undergoing coronary artery surgery. J Cardiothor Vasc Anesth
12. Ben-Haim SA, Amar R, Shofty R, Dinnar U. Low positive end expiratory pressures improve the left ventricular workload vs. coronary blood flow relationship. J Cardiovasc Surg
13. Roissant R, Slama K, Jaeger M et al.
Fluid restriction and early extubation for successful liver transplantation
. Transplant Proc
14. Kaisers U, Langrehr JM, Haack M et al.
Hepatic venous catheterization in patients undergoing positive end-expiratory pressure ventilation after OLT: technique and clinical impact. Clin Transplant
15. Mandell MS, Lockrem J, Kelley SD. Immediate tracheal extubation after liver transplantation
: experience of two transplant centers. Anesth Analg
16. Neelakanta G, Sopher M, Chan S et al.
Early extubation after liver transplantation
: early extubation is not preferable to late extubation in patients undergoing coronary artery surgery. J Cardiothor Vasc Anesth
17. Glanemann M, Langrehr J, Kaisers U et al.
Postoperative tracheal extubation after orthotopic liver transplantation
. Acta Anaesthesiol Scand
18. Findlay JY, Jankowski CJ, Vasdev G et al.
Fast track anaesthesia for liver transplantation
reduces postoperative ventilation time but not intensive care unit stay. Liver Transpl
19. Lindop M. We have a liver. Have we got a bed? Intensive care support for liver transplantation
surgery. Liver Transpl