As a result of the increasing incidence of cirrhosis in the UK, more patients with chronic liver disease are being considered for elective non-hepatic surgery. A historical reluctance to offer surgery to such patients stems from general perceptions of poor postoperative outcomes. While this is true for those with decompensated cirrhosis, selected patients with compensated early-stage cirrhosis can have good outcomes after careful risk assessment. Well-recognised risks include those of general anaesthesia, bleeding, infections, impaired wound healing, acute kidney injury and cardiovascular compromise. Intra-abdominal or cardiothoracic surgery are particularly high-risk interventions. Clinical assessment supplemented by blood tests, imaging, liver stiffness measurement, endoscopy and assessment of portal pressure (derived from the hepatic venous pressure gradient) can facilitate risk stratification. Traditional prognostic scoring systems including the Child-Turcotte-Pugh and Model for End-stage Liver Disease are helpful but may overestimate surgical risk. Specific prognostic scores like Mayo Risk Score, VOCAL-Penn and ADOPT-LC can add precision to risk assessment. Measures to mitigate risk include careful management of varices, nutritional optimisation and where possible addressing any ongoing aetiological drivers such as alcohol consumption. The role of portal decompression such as transjugular intrahepatic portosystemic shunting can be considered in selected high-risk patients, but further prospective study of this approach is required. It is of paramount importance that patients are discussed in a multidisciplinary forum, and that patients are carefully counselled about potential risks and benefits.
- ABDOMINAL SURGERY
- PORTAL HYPERTENSION
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
The increasing incidence of cirrhosis in the UK1 has driven the rise in such patients requiring non-hepatic surgery as an emergency (eg, irreducible umbilical hernia) or electively (eg, cardiac or colon cancer surgery). There has been a long-standing reticence to operate on such patients due to their perceived poor outcomes, which has potentially disadvantaged this patient group.
Surgical outcomes vary considerably according to the type of surgery, type of anaesthesia and importantly the stage of liver disease, with clinically significant portal hypertension ((CSPH) defined as a hepatic venous pressure gradient (HVPG) ≥10 mm Hg) heralding the development of varices, and further progression to decompensation.2 Additionally, comorbidities such as obesity, diabetes and cardiovascular disease present further anaesthetic and surgical challenges for the growing cohort of patients with non-alcoholic fatty liver disease-related cirrhosis.
The assessment of surgical risk in patients with cirrhosis is increasingly important but limited by a lack of prospective controlled data and significant heterogeneity in risk assessment pathways. Although AGA guidance on this topic has been published,3 there is no clear guidance in the UK. This position statement aims to summarise the current evidence and advise members of BSG and BASL on the current options for presurgical assessment and future directions. This guidance document has been reviewed by the BASL Portal Hypertension Special Interest Group Steering Committee, BSG Liver Section, and BSG Clinical Services and Standards Committee.
Patients with cirrhosis are considered high-risk candidates for non-hepatic surgery, in particular intra-abdominal and cardiothoracic procedures. The risk is highest inChild-Turcotte-Pugh (CTP)-B and C disease and correlates with the degree of portal hypertension.
The preoperative assessment of patients with cirrhosis being considered for non-hepatic surgery is complex and a case-by-case decision is advised.
Multidisciplinary assessment including hepatology, anaesthetics and surgery is advised, with involvement of other disciplines (eg, radiology, haematology or nutrition/physiotherapy) where appropriate.
In eligible patients, particularly those with decompensated cirrhosis, consideration should be given to deferring elective surgery until after liver transplantation where applicable. Where liver transplantation is not an option in decompensated cirrhosis, referral to palliative care services should be considered unless aetiology of cirrhosis is modifiable for example, alcohol abstinence, antiviral therapy.
The current bespoke prognostic scoring systems such as Mayo Risk Score, VOCAL-Penn and Adequate Operative Treatment for Liver Cirrhosis (ADOPT-LC) can be considered to aid multidisciplinary team decisions, in addition to CTP and Model for End-stage Liver Disease (MELD) score.
Measuring HVPG can provide additional important information and may be considered as an adjunct to other prognostic paraments where available.
In compensated cirrhosis, transjugular intrahepatic portosystemic shunting (TIPSS) prior to surgery can be considered in those at high risk as quantified by the presence of varices, prognostics scoring systems and/or HVPG. However, further comparative multicentre studies are urgently needed, to guide selection of patient most likely to benefit
Algorithm for assessment of patients with cirrhosis undergoing elective surgery is presented in figure 1.
Presurgery evaluation requires that we consider specific harms that may either complicate surgery or recovery from surgery. These risks should be assessed and explained to the patient, with appropriate steps put in place to mitigate where possible.
History, physical examination and blood tests may reveal findings suggestive of chronic liver disease. There should be a low threshold to consider further investigations such as imaging and liver stiffness measurement where there is clinical suspicion of liver disease since patients with compensated advanced chronic liver disease can present with normal liver function tests and absence of overt clinical stigmata of chronic liver disease. A multidisciplinary assessment is strongly advised with input from hepatology, anaesthetics, surgery and other disciplines where appropriate such as radiology, haematology, nutrition and physiotherapy. The risks include: (1) general anaesthesia which could be mitigated by the use of regional techniques; (2) implied bleeding risk using traditional indices such as international normalised ratio (INR), activated partial thromboplastin time (APTT) and platelet count, which in this population do not reflect deranged haemostatic mechanisms, making an assessment of bleeding risk challenging4; (3) infection which can precipitate decompensation5; (4) compromised nutritional status with malnutrition and sarcopenia hampering postoperative recovery and wound healing6; (5) acute kidney injury precipitated by hypovolaemia due to depleted albumin, contrast agents, infection or anaesthesia6; (6) cardiovascular compromise due to stress of surgery unmasking undiagnosed cardiomyopathy.7 8 Decompensated cirrhosis (defined by ascites, encephalopathy, variceal bleeding, and/or jaundice) results in a much higher risk of poor outcomes, with increased propensity to precipitate acute on chronic liver failure postsurgery.9 In this group, an individualised decision on whether surgery can be deferred until after liver transplantation for eligible candidates should be considered. Decompensated patients with non-hepatic malignancy can be particularly challenging since pre-emptive liver transplantation is normally contraindicated. Elective surgery is absolutely contraindicated in acute liver failure, acute viral or alcoholic hepatitis and American Society of Anesthesiologists (ASA) Physical Status class V. Intra-abdominal or cardiothoracic surgical procedures are considered higher risk—the former due to the potential presence of portal hypertension, and the latter due to the adverse impact of reduced cardiac output in patients who are already vasodilated, with low blood pressure and reliant on their compensatory high cardiac output state.10
Specific prognostic models
Most studies of prognostic models are observational or retrospective and limited by significant heterogeneity with regard to the type of surgery, severity of liver disease and whether surgery was elective or emergency. Traditional prognostic scoring systems including CTP and MELD, have an established role but can overestimate surgical risk.11 Bespoke prognostic models have also been proposed which can provide a greater precision (table 1). These include the Mayo Risk Score,12 VOCAL-Penn13 and ADOPT-LC score14 (table 1). These are all based on retrospective data, some dating back 25 years, and do not quantify the degree of portal hypertension. A recent multicentre prospective study15 found that ASA class (HR III vs II=2.98 (95% CI 0.7 to 13.2), HR IV vs II=9.97 (95% CI 2.0 to 50.4), p=0.008), high-risk surgery (HR=3.65 (95% CI 1.4 to 9.3), p=0.006) and HVPG (HR=1.14 (95% CI 1.05 to 1.25), p=0.003) were strongly predictive of 1 year mortality (C-statistic >0.8). HVPG is invasive and not widely available, but given that values >16 mm Hg (and especially ≥20 mm Hg (HR=5.67 (95% CI 2.4 to 13.2)) are independently associated with high risk of 1-year postsurgical mortality, its use, particularly in patients with compensated cirrhosis or with previous decompensation, may be considered where available.
Risk associated with different types of non-hepatic surgery
Patients with liver disease have high morbidity and mortality across the spectrum of surgical procedures.16 Operative risk is dependent on severity of liver disease, type and mode of surgical procedure, concomitant medical conditions, nutritional and performance status of the individual, and peri- and postoperative expertise.17 18 Studies on the outcome of patients undergoing surgery are highly variable with the majority of studies being retrospective and single-centre, and lacking detail of liver disease and appropriate quality controls17 19–22 (table 2). Emergency surgery confers an estimated 4–10 fold higher postoperative mortality.10 23–26 A systematic review by de Goede et al quotes general surgical risk for 30-day mortality and morbidity to be 11.6% and 30.1%, respectively.27 Presence of portal hypertension further heightens the mortality with colectomy, abdominal aortic aneurysm repair or coronary artery bypass grafting (CABG) having a 12.3-fold, 14.3-fold, 7.8-fold and 22.7-fold higher risk, respectively, when compared with patients without cirrhosis.27
Laparoscopic versus open surgery
Laparoscopic surgery was deemed a relative contraindication in cirrhosis due to risk of rupturing abdominal varices but multiple studies have demonstrated favourable outcomes following this approach.28–37 It’s safety in CTP-C patients is not yet proven, hence surgery in such patients may be limited to conservative measures. In the context of cirrhosis, laparoscopic cholecystectomy has been the most common procedure assessed. In a meta-analysis of randomised controlled trials comparing laparoscopic and open cholecystectomy (n=234 patients (97% CTP-A or B)), there were no postoperative deaths and the laparoscopic approach was associated with fewer postoperative complications (p=0.03), shorter hospital stay (p<0.001) and quicker resumption of normal diet (p<0.001).29 Laparoscopic inguinal hernia repair can be performed safely32 and there are also favourable outcomes when comparing laparoscopic to open appendectomy.38
Colorectal surgery in cirrhosis is considered high risk with morbidity and 30-day mortality quoted between 21.5%–26% and 48%–77%, respectively.39–43 The risk is higher in the presence of portal hypertension due to altered intraoperative haemodynamics, ascites and coagulopathy.10 40 Ascites heightens the risk of wound complications such as dehiscence.44 A meta-analysis demonstrated that, compared with a non-cirrhotic group, patients with cirrhosis who underwent primary colorectal cancer surgery had more major complications, a higher reoperation rate and increased short-term mortality45 (table 2). Additional risk factors for mortality included older age, elevated bilirubin, prolonged prothrombin time, higher CTP class, MELD score >15, intraoperative transfusions and comorbidities (cardiovascular disease, chronic kidney disease, paraplegia and malnutrition).39 46–48 Nonetheless, improvements in surgical technique and perioperative care have led to reduced mortality rates over the last two decades. Mortality was reported to be 13% in a study of 72 cirrhosis patients undergoing colorectal surgery in 2003.46 However, in a more recent study, Lee et al reported mortality of 3.1% 161 patients undergoing colorectal surgery.48
The incidence of abdominal wall hernia in patients with cirrhosis is 16%, which rises to 24% in patients with ascites.49 Half of these are umbilical hernias and 60% of patients experience recurrence following repair.49 Mortality is quoted to be 6% for umbilical hernia repair, which is higher in the emergency setting.50 However, a randomised controlled trial concluded that elective repair of an umbilical hernia is safe in cirrhosis patients, even in patients with a relatively high MELD score (table 2). The presence of ascites appears to be strongly predictive of hernia recurrence.51 52
Morbidity and mortality in patients with cirrhosis undergoing cardiovascular surgery is increased with higher risk of infectious, respiratory and renal complications alongside longer hospital stay.53–56 CABG and valvular surgery carry approximately fivefold and sevenfold increased risk of mortality in cirrhosis, respectively.56 A meta-analysis (CABG, valvular surgery and cardiopulmonary bypass) quoted 1-year mortality at 20.6%, 43.6% and 56.5% for patients with CTP A, B and C, respectively.57 Both CTP and MELD score reliably predict mortality,58 and cardiopulmonary bypass is not recommended in patients with an MELD >13.5 or CTP >8.59 60 Despite these data, cardiac surgery risk prediction models do not consider liver dysfunction as a surgical risk factor.61 Surgical aortic valve replacement carries higher mortality in comparison to transcatheter aortic valve replacement (TAVR), therefore, TAVR is recommended in cirrhotic patients.62 There are limited data on abdominal aortic aneurysm repair but CTP-B and MELD >10 are associated with reduced survival.63
Upper gastrointestinal and bariatric surgery
Upper gastrointestinal surgery outcomes in patients with cirrhosis are variable. Emergency surgery for complicated peptic ulcer disease (perforation and bleeding) is associated with a high mortality rate (23%–64%).64 For gastric cancer operations, morbidity and mortality in cirrhosis patients is 25.6% and 10.3%, respectively.65 Surgery carries acceptable risks for CTP-A and CTP-B cirrhotic patients; thus gastrectomy with D2 or more lymph node dissection can be safely carried out in CTP-A patients, whereas only D1 lymph node dissection is recommended in CTP-B patients. Radical gastrectomy is likely to be fatal in CTP-C patients, with mortality rates of 100% reported.66 67
Oesophagectomy in cirrhosis patients carries a high risk of developing pulmonary complications (postoperative pneumonia, pleural effusions and chylothorax), ascites, anastomotic leaks during the first month, in addition to longer intensive care stays in comparison to patients without cirrhosis.68 69 Patients who are CTP-A have significantly lower mortality in comparison to CTP-B patients, and CTP-C disease remains a contraindication to oesophagectomy.68 MELD score >9 was associated with significantly lower 5-year survival and a score ≤9 was associated with similar outcomes to non-cirrhotic controls.70
Bariatric surgery is considered safe to be undertaken in an experienced centre. A meta-analysis and systematic review concluded that postoperative and liver-related complications were higher among patients with cirrhosis when compared with non-cirrhotic individuals. Significantly lower postoperative complications were noted with sleeve gastrectomy compared with Roux-en-Y gastric bypass.71
The lifetime risk of appendicitis is 8.6% in males and 6.7% in females, making it the most common surgical emergency worldwide.72 A meta-analysis showed that 30-day mortality for cirrhosis patients undergoing appendicectomy was 9%, in comparison to 0.3% in those without cirrhosis. In patients with compensated cirrhosis, laparoscopic appendicectomy is safer compared with an open approach. However, in many studies there was a lack of information on the severity of cirrhosis and other patient characteristics.73
Adverse events are reported for cirrhosis patients undergoing total knee and hip arthroplasty but studies are mostly retrospective and limited.74–77 In a systematic review, cirrhotic patients are more likely to experience postoperative haemorrhage, surgical site infection and need for revision surgery after total hip arthroplasty.78 However, the severity of liver disease was not presented, and some studies report only short-term outcomes (table 2).
Interventions to mitigate the risk of surgery in patients with cirrhosis
A joint care approach is desirable to mimise the additional risk of surgery in patients with cirrhosis. The degree of portal hypertension can accurately predict postsurgical outcomes in cirrhosis. In the study by Reverter et al, patients with HVPG >16 mm Hg and ≥20 mm Hg were considered high and very high risk, respectively. Liver stiffness measurement of >25 kPa also correlates with CSPH and can be helpful in stratifying compensated cirrhosis.79 Endoscopy and contrast-enhanced CT scanning of the abdomen are indicated in all patients with cirrhosis. National guidance for endoscopic and pharmacological management of varices should be followed.80 81 Attention should also be given to optimising nutrition where possible.6 An area of debate is the role of portal decompression prior to surgery. Prophylactic TIPSS may facilitate planned surgery by reducing complications pertaining to portal hypertension. Uncontrolled retrospective studies suggest a potential role for TIPSS prior to surgery in selected patients82–86 (table 3). However, there is no evidence of improved surgical outcomes, and patient selection is not defined, although HVPG could be helpful as a guide. Moreover, the optimal timing of surgery after TIPSS is unclear and procedure-related complications should be considered. The present BSG guidance advises further research.81
Areas requiring further study
Role of HVPG in stratifying risk prior to surgery, in patients with compensated cirrhosis or prior decompensation. Controlled studies are recommended.
Further multicentre prospective study on the utility of bespoke prognostic scoring systems. For patients with decompensated cirrhosis, these might include the EF-Chronic Liver Failure Acute Decompensation Score to guide risk.
In compensated cirrhosis, assessment of non-invasive markers of CSPH, such as liver stiffness measurements as a means of stratifying risk and selection for interventions prior to surgery.
The role of facilitative TIPSS prior to surgery. Further prospective comparative multicentre studies with a focus on patient selection criteria, for example, HVPG or liver stiffness measurements and postoperative outcomes.
Prospective studies to address the role of portal pressure lowering agents prior to surgery to reduce risk, including carvedilol, statins and combinations including rifaximin.
Prospective studies to investigate any potential role for prehabilitation within these cohorts of patients, which may improve surgical outcomes and reduce postoperative complications.
Further study to investigate the role of dynamic function testing such as viscoelastic tests and newer assays such as platelet function assays in predicting bleeding risk.
Patient consent for publication
Twitter @Nadirabbas_x, @NeilRaj1
Correction notice This article has been corrected since it published Online First. A typographical error has been corrected in the abstract.
Contributors NA and DT contributed to the conception and design of guidelines documented and wrote the first draft. All coauthors contributed to critical revision of the article.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not–for–profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.