Introduction

“Fast-track” or “enhanced recovery after surgery” (ERAS) programs following surgical interventions is now standard in the care of patients with several surgical indications [1–5]. These programs use a multimodal approach to maximize effectiveness and minimize cost, thus optimizing perioperative care pathways [6]. ERAS programs have been associated with reduced complication rates, hospital stays, and hospital costs [1]. However, there is limited evidence of the effectiveness of ERAS programs for patients undergoing liver surgery, with just six small published series [7]. Among these, two small randomized trials suggest that ERAS following hepatectomy is associated with shorter hospitalization and reduced morbidity [8, 9]. Most of the patients in these studies underwent hepatectomy for colorectal liver metastases; however, a few studies have analyzed the effect of ERAS on patients who underwent hepatectomy for hepatocellular carcinoma (HCC).

HCC is the fifth most common cancer in the world [10]. Although most affected patients are in Asia and Africa, the incidence of HCC and its mortality rates are increasing in North America and Europe [11, 12]. In Japan, most HCCs develop in patients with chronic hepatitis and liver cirrhosis induced by hepatitis B or C virus infection. With advances in perioperative management, anesthesia, and operative techniques, hepatectomy for HCC is being performed more commonly [13]. However, the postoperative mortality rate of these patients remains higher than that of patients with cirrhosis or chronic hepatitis undergoing other types of surgery. The morbidity rate of patients with cirrhosis undergoing liver resection has been reported to the range from 20 to 70 %, with mortality rates of 5–21 % [14–19]. However, mortality rates at high-volume centers in Japan are usually much lower, at <2 % [20–22], although morbidity rates remain relatively high. The postoperative course of these patients does not always proceed as expected, which is attributed to various forms of intraoperative stress, including blood loss and ischemia. These findings emphasize the importance of improving both surgical techniques and perioperative care to reduce the mortality and morbidity rates of HCC patients undergoing liver resection.

We hypothesized that the application of an ERAS program for HCC patients undergoing hepatectomy would accelerate recovery, reduce morbidity rates, and shorten the hospital stay. Thus, we conducted this study to assess whether an ERAS program for HCC patients affected the feasibility, safety, and effectiveness of potentially curative hepatic resection.

Materials and methods

Patients

Between January, 2008 and December, 2013, 315 patients underwent R0 resection, defined as macroscopic removal of all tumors, for HCC at our institution. The 130 patients, who underwent resection between January 2008 and December 2010, 41 patients, who underwent resection between January 2011 and January 2012, and 144 patients, who underwent resection between February 2012 and December 2013, were included in this number. Eight patients died in hospital after surgery: four, one, and three patients, respectively, from each of the three time periods. Data on all patients were recorded prospectively. The study protocol was explained to all patients, so that those who underwent surgery between January 2008 and December 2010 understood that they would receive the conventional perioperative management and those who underwent surgery between February 2012 and December 2013 understood that they would be managed using the ERAS program. The study protocol was approved by the institutional ethics committee of our institution.

Patients were subclassified based on whether they underwent removal of two or more sections. The 71 patients who underwent removal of more than two sections were then subclassified into an ERAS group (n = 47) and a control group (n = 24).

Clinicopathologic variables and surgery

Before surgery, each patient underwent conventional liver function tests and measurement of the indocyanine green (ICG) retention rate at 15 min. Patients were screened for hepatitis by measuring serum concentrations of hepatitis B surface antigen and antibody to hepatitis C virus. Serum concentrations of α-fetoprotein and protein induced by vitamin K absence/antagonism-II were also measured in all patients. Surgical procedures were classified according to the Brisbane terminology [23]. Anatomic resection was defined as the resection of the tumor together with the related portal vein branches and the corresponding hepatic territory. Anatomic resection was subclassified as hemihepatectomy (resection of half the liver), extended hemihepatectomy (hemihepatectomy plus removal of additional contiguous segments), sectionectomy (resection of two Couinaud subsegments [24]), or segmentectomy (resection of one Couinaud subsegment). All other non-anatomic procedures were classified as limited resection. The tumors treated with limited resection included both peripheral and central tumors. Patients with peripheral tumors and those with extrahepatic growth underwent partial hepatectomy, because this method achieved adequate surgical margins. In contrast, patients with central tumors located near the hepatic hilum or major vessels underwent enucleation because of the difficulty and risks associated with obtaining adequate margins. In this study, the resection of more than two sections was defined as extended hepatectomy. One senior pathologist reviewed each specimen for histologic confirmation of the diagnosis. Perioperative/postoperative complications and deaths were recorded to assess the morbidity and mortality associated with hepatectomy. Complications were classified by the primary investigator, according to the Clavien–Dindo classification scale [25, 26]. Grade I and II complications were classified as minor morbidities, and grade III and IV complications were classified as major morbidities. Grade V was defined as patient death.

Liver resections in both the ERAS and control groups were performed by the same surgeons. Right subcostal abdominal incisions were used. The transection plane was determined by intraoperative ultrasonography. The Cavitro Ultrasonic Surgical Aspiration (CUSA®; Valleylab, Boulder, CO, USA) and Aquamantys® bipolar (Medtronic, Minneapolis, MN, USA) were used for liver parenchymal dissection.

No specific measures were taken to avoid prolonged preoperative and postoperative fasting, nasogastric decompression, excessive use of intravenous fluids, or prophylactic abdominal drains. The conventional postoperative care program emphasized prolonged rest for both the patient and the gastrointestinal tract. The ERAS multimodal evidence-based recovery program protocol, which was designed originally for elective colonic surgery, was modified to cover all aspects of elective liver resection [27]. Table 1 lists the details of the ERAS liver program.

Table 1 Care plan for patients undergoing liver resection in the “enhanced recovery after surgery” (ERAS) program
Full size table

Perioperative and postoperative exercises by HCC patients with hepatic impairment led to weight loss, as a result of reductions in fat mass and improved insulin resistance, but had no effect on skeletal muscle mass. Intensifying perioperative and postoperative exercise may better maintain postoperative physical strength and result in earlier resumption of daily activities. An exercise program was tailored for each patient. Exercise was started as soon as possible after diagnosis, up to 1 month preoperatively, and was resumed 1 week postoperatively, and then continued for 6 months. The program consisted of three 60-min exercise sessions per week. Each session included 5 min of stretching exercises, 30 min of walking at an intensity based on the anaerobic threshold of each patient, 20 min of targeted stretching exercises, and 5 min of cooling down with stretching. Once or twice a month postoperatively, a medical doctor and exercise trainer confirmed the frequency and quantity of exercise performed by each patient. Patients were advised to continue exercising for 6 months after surgery [28]. Following conclusion of this randomized control trial, patients with cirrhosis and HCC in our institution were prescribed perioperative exercise therapy, together with the administration of branched chain amino acids (BCAA).

Patients were instructed to drink up to 1000–1500 mL of carbohydrates (OS-1®; carbohydrate 2.5 g/100 mL, glucose 1.8 g/100 mL), from the evening of the day prior to surgery until 3 h before surgery. Although no formal discharge criteria were specified for patients in the control group, discharge criteria were specified in the ERAS group, and included the following: normal or decreasing serum bilirubin and normal or increasing serum albumin concentrations; good pain control with oral analgesia only; tolerance of solid food; no intravenous fluids; mobile independently or at the preoperative level; and willingness to go home.

Statistical analysis

Continuous variables are presented as medians and the first and third quartiles. Differences between groups were assessed by the Chi-square test or the Wilcoxon rank sum test, as appropriate. Differences between groups in the perioperative concentrations of albumin, cholinesterase, and C-reactive protein were assessed by a two-way analysis of variance. The level of significance was set at P < 0.05. The Kaplan–Meier life table method was used to calculate the probability of hospitalization beyond specific time-points, with differences estimated using the generalized log-rank test. All statistical analyses were performed with SPSS® for Windows 11.0J (SPSS Inc., Chicago, IL, USA).

Results

Differences between the ERAS and control groups

Table 2 summarizes the differences between the ERAS and control groups. The ability to commence exercise therapy with BCAA within 1 month before surgery and the start of exercise therapy with BCAA on postoperative days 4–6 were limited for patients with cirrhosis; therefore, these criteria were achieved by 41 and 46 % of the ERAS group patients before and after surgery, respectively. Over 87 % of the patients in the ERAS group ate normally until midnight on the day before surgery, they did not undergo preoperative bowel preparation, and they received carbohydrate drinks until 3 h before surgery. In contrast, patients in the control group started fasting after the evening meal and refrained from drinking fluids after midnight on the day before surgery. Intraoperatively, drainage of the peritoneal cavity was omitted for 60 % of the patients in the ERAS group. The nasogastric tube was removed immediately after surgery in 137 of the 144 patients in the ERAS group, and the next day in the remaining seven patients because of slower awakening from anesthesia. Oral intake of rice gruel commencing from the evening meal and patient mobilization at least three times per day were achieved by 98 (68 %) and 118 (82 %) patients, respectively, in the ERAS group. Urinary catheters were removed, and intravenous fluids were discontinued on median postoperative days 5 and 6, respectively, in the control group. By postoperative day 3, however, 90 % of patients in the ERAS group have hadtheir catheter removed (median 3 days) and 67 % had discontinued intravenous fluids (median 3 days).

Table 2 Differences between the “enhanced recovery after surgery” (ERAS) and control groups
Full size table

Preoperative characteristics

Table 3 summarizes the preoperative characteristics of the HCC patients in the control and ERAS groups, who underwent removal of more than two segments of the liver. There were no between-group differences in sex, age, hepatitis virus status, preoperative liver function (ICG retention rate, serum albumin, total bilirubin, alanine aminotransferase, and platelet count), C-reactive protein, α-fetoprotein, or protein induced by vitamin K absence/antagonism-II.

Table 3 Preoperative characteristics of hepatocellular carcinoma (HCC) patients in the “enhanced recovery after surgery” (ERAS) and control groups, who underwent two or more sectionectomies
Full size table

Perioperative parameters and pathologic findings

As shown in Table 4, the procedures used for anatomic resection, blood loss, blood transfusion, postoperative complications, and hospital death rate did not differ significantly between the control and ERAS groups. However, the percentage of patients with a severe grade on the Clavien–Dindo classification was higher in the control group than in the ERAS group. Operating times and postoperative hospital stay were significantly shorter, and the total volume of blood infused during surgery was significantly lower in the ERAS than in the control group. The median postoperative hospital stay was 13.0 days in the ERAS group and 16.5 days in the control group. The probability of hospitalization beyond a given time point was significantly lower in the ERAS group than in the control group (Fig. 1). In the control group, two patients had grade IVa complications on the Clavien–Dindo classification and were discharged from hospital on postoperative days 111 and 147, respectively. The patient who required a postoperative hospital stay of 111 days underwent right hemihepatectomy for HCV-associated HCC and was treated postoperatively for intractable ascites, pleural effusion, and aspirational pneumonia. This patient had a preoperative serum albumin concentration of 3.2 g/dL and serum albumin concentrations immediately after surgery of 2.0–2.5 mg/dL. BCAA supplementation resulted in recovery of the preoperative serum albumin concentration 3 months after surgery. The second patient underwent abdominal drain placement during surgery, but bile leakage through the drain appeared on postoperative day 2. Despite performing an intraoperative bile leakage test with saline solution, the site of major leakage could not be detected. This patient required a postoperative hospital stay of 147 days to treat postoperative bile leakage, which proved extremely difficult to resolve.

Table 4 Operative and postoperative characteristics of hepatocellular carcinoma (HCC) patients in the “enhanced recovery after surgery” (ERAS) and control groups who underwent two or more sectionectomies
Full size table
Fig. 1
figure 1

Probability of hospitalization after resection of hepatocellular carcinoma (HCC) in patients in the control (dotted line) and “enhanced recovery after surgery” (ERAS) (unbroken line) groups. The probability of hospitalization differed significantly (P = 0.033). The numbers of patients at risk are shown below the graph

Full size image

Although the number of patients with retention of abdominal drainage was significantly lower in the ERAS group, the frequency of abdominal paracentesis in patients without intraoperative abdominal drainage was higher in this group. Abdominal paracentesis was caused by refractory ascites in five patients, by refractory ascites and pleural effusion in one patient, and by intra-abdominal abscesses in two patients. Although the difference was not significant, the median operative blood loss was greater (1636 vs 1154 mL), and the median operating time was longer (488 vs. 393 min) in the eight patients with abdominal paracentesis than in the 16 patients without intraoperative abdominal drainage or postoperative paracentesis.

A comparison of the pathological findings in the two groups showed no differences in tumor size, number of tumors per patient, associated liver diseases, or TNM stage (Table 4).

Physical and biochemical parameters

The percentage of patients who resumed dietary intake on postoperative day 2 was significantly higher in the ERAS than in the control group (Fig. 2). Although the time to first bowel movement did not differ significantly between the groups, the ability to walk steadily occurred significantly earlier in the ERAS group than in the control group (Fig. 3). Postoperative serum concentrations of albumin and cholinesterase were significantly higher in the ERAS group than in the control group (Fig. 4). C-reactive protein concentration was lower in the ERAS group, but the difference was not significant.

Fig. 2
figure 2

Times until dietary intake after resection of hepatocellular carcinoma (HCC) in patients in the control (open squares) and “enhanced recovery after surgery” (ERAS) (closed squares) groups. Dietary intake by postoperative day 2 was significantly more frequent (P = 0.049) in the ERAS group than in the control group

Full size image
Fig. 3
figure 3

Times to ability to walk steadily and to first bowel movement after resection of hepatocellular carcinoma (HCC) among patients in the control (open squares) and “enhanced recovery after surgery” (ERAS) (closed squares) groups. The ERAS group patients were able to walk steadily significantly earlier (P = 0.031) in the ERAS than the control group patients

Full size image
Fig. 4
figure 4

Serum concentration of a albumin, b cholinesterase, and c C-reactive protein in the “enhanced recovery after surgery” (ERAS) (closed circles) and control (open circles) groups after liver resection. Data are expressed as the mean ± standard deviation

Full size image

Discussion

The ERAS protocol did not compromise the safety of HCC patients with chronic liver disease, undergoing liver resection. This study showed that using an evidence-based multimodal enhanced program following the removal of more than two liver segments for HCC accelerated postoperative recovery and reduced the hospital stay significantly. Although postoperative recovery did not differ significantly between the two patients groups overall, it differed significantly in patients who underwent extended surgery, suggesting that the effectiveness of ERAS was limited.

Early postoperative enteral nutrition has been found to improve clinical outcomes, compared with ‘nil by mouth’ [29]. In this study, patients managed according to the ERAS protocol were able to drink fluids and eat normal food on the day of surgery. Early resumption of normal diet in combination with other elements of the enhanced recovery program is designed to reduce delayed gastrointestinal function after surgery and can even promote appetite. A restrictive perioperative intravenous fluid regimen may also help prevent delayed gastrointestinal functioning [30]. The total volume of blood infused during the operation was significantly less in the ERAS group than in the control group. Fluid restriction may be important in hepatic surgery. In this study, postoperative dietary intake was significantly higher in the ERAS group.

It is important to treat surgical patients in an environment that encourages early mobilization [31]. In this study, mobilization was not achieved on the day of surgery in any HCC patient who underwent removal of more than two liver segments. Adequate pain control and a substantial effort by the nursing staff are required for early mobilization. Patients in the ERAS group were completely mobile after a median 5.5 days, and their median total hospital stay was 13.0 vs 16.5 days in the control group. Two of the control group patients suffered complications classified as grade IVa on the Clavien–Dindo scale and were discharged from hospital on postoperative days 111 and 147, respectively. Excluding these two patients from the control group would eliminate the significant difference between the groups (data not shown; P = 0.052). It is unclear whether the absence of the ERAS program was associated with the development of postoperative complications in these two patients, one of them whom suffered intractable ascites and the other, bile leakage. Although the patient with intractable ascites did not receive perioperative BCAA, its postoperative administration during month 2 restored his albumin concentration to the preoperative level and reduced the volume of ascites. In contrast, none of the patients in the ERAS group who received perioperative BCAA suffered postoperative intractable ascites after starting exercise therapy.

We have attempted to eliminate the retention of abdominal drains after starting ERAS, by performing intraoperative ICG fluorescent cholangiography. It is our practice to insert abdominal drains after liver surgery to monitor postoperative bleeding, reduce abdominal pressure caused by intractable ascites, and detect postoperative bile leakage and drainage. We have found that intraoperative ICG fluorescent cholangiography can detect insufficiently closed bile duct stumps not detected by standard bile leak tests [32]. Although two patients in our ERAS group experienced postoperative intra-abdominal abscesses caused by minor bile leakage, both were started on antibiotics within 2 weeks after surgery.

Japan has a universal public health insurance system, which covers 70–90 % of all medical costs, making hospital stays very affordable. Consequently, early discharge is not a priority, and some patients are reluctant to accept a shorter hospital stay. Short-term outcomes have been reported in 7732 patients who underwent hepatectomy for more than one segment during 2011 in 987 different hospitals, as identified in the National Clinical Database of Japan [33]. In fact, the mean and median hospital stays after HCC surgery were 23.7 and 16.0 days, respectively.

Almost all of the patients in this study had liver dysfunction, such as chronic hepatitis or cirrhosis. Patients with liver dysfunction started exercise therapy with BCAA within 1 month before surgery and restarted exercise therapy with BCAA on postoperative days 4–6. Among patients who underwent removal of two or more sections, 22 of 47 in the ERAS group and 10 of 24 in the control group started exercise therapy and received BCAA. Because the fasting time was shorter in the ERAS group, their increased postoperative dietary intake and perioperative exercise therapy with BCAA may have resulted in higher serum concentrations of albumin and cholinesterase, which are markers of nutritional status, during the early postoperative period.

It is unclear if the ERAS program reduces morbidity rates after liver resection. Randomized trials performed in the UK [8] and China [9] showed lower complication rates in ERAS patients, but other studies have failed to show any differences. In this study, the overall complication rates before and after surgery were similar in the ERAS and control groups. However, severe complications (Clavien–Dindo grade IV) were less common in the ERAS group.

Prophylactic drains are used by many centers to detect early complications, such as postoperative hemorrhage and bile leakage, to remove intraperitoneal fluids, and to prevent abscess formation. However, abdominal drainage after liver resection may not reduce the incidence of reinterventions needed for postoperative complications [34–36]. Several studies have reported higher rates of infected collections when drainage was used [34, 37, 38], suggesting that drains may be detrimental to clinical outcomes by providing a route for ascending infections. Another disadvantage of drains in enhanced recovery settings is that they represent a significant impediment to achieving early mobilization. In this study, although the number of patients without intraoperative abdominal drains was significantly higher in the ERAS group than in the control group, the frequency of abdominal paracentesis was higher in the ERAS group. Abdominal paracentesis is frequently caused by refractory ascites. Postoperative abdominal paracentesis in these patients may have been avoided by the installation of abdominal drains. One patient in the control group required a postoperative hospital stay of 147 days to treat postoperative bile leakage. As the insertion of an abdominal drain may prevent complications after major bile leakage, omitting prophylactic intraoperative installation of abdominal drains in HCC patients with liver dysfunction, as part of the ERAS program, seems disadvantageous.

The period from January 2011 to January 2012 was considered a transitional period, so patients who underwent surgery during this period were omitted from the two groups as any of the 16 items in the ERAS program may or may not have been performed by attending surgeons. However, from February 2012, all attending surgeons essentially performed all of the items in the ERAS program. Although these interventions were regarded as minor changes from the conventional perioperative management, our results suggest that minor changes in the ERAS program resulted in some alterations in postoperative outcomes.

This study had several limitations. Although the study was not randomized, data were collected prospectively, with the principal aim being to assess the impact of implementing an enhanced recovery pathway in routine clinical practice. A randomized clinical trial comparing enhanced recovery pathways and traditional care was considered impractical. Based on increasing evidence of the efficacy of ERAS programs, we implemented ERAS protocols for almost 1 year, regarding the traditional perioperative protocol as suboptimal care. More importantly, it was necessary for nursing and surgical staff to start providing enhanced recovery programs. Therefore, this feasibility study was designed as a prospective care study with a historical control group.

In conclusion, a multimodal enhanced recovery program proved feasible and effective for patients with chronic liver diseases undergoing extended liver resection for HCC. Patients were able to drink and eat the day following surgery and most were mobile by postoperative day 6. Additional studies are warranted to confirm the effectiveness of this protocol, as determined by patient recovery, and to determine whether to include other components of the ERAS program.