Delayed Return of Gastrointestinal Function after Partial Hepatectomy: A Single-Center Cross-Sectional Study Open Access

Partial hepatectomy (PH) is the standard treatment for several liver diseases, including malignant and benign tumors. Despite improvement in perioperative management, technological advances, and extensive experience of specialist centers, liver surgery still is a surgery with a high morbidity rate reaching 30–50% [1‒5].

Certain complications such as postoperative hemorrhage, biliary leakage or liver failure have been extensively studied, and their risk factors are well-known [6‒11]. Conversely, delayed return of gastrointestinal function (DRGF) after PH remains unclear. This entity encapsulates both paralytic ileus (PI) and delayed gastric emptying (DGE). The former refers to the absence of normal peristalsis of the gastrointestinal tract after surgery whereas the latter stands for conditions in which there is a decrease in gastric motility. While PI and DGE are typical postoperative complications of colorectal and pancreatic surgery, respectively, their diagnosis is much more challenging after PH; they show similar clinical presentations and require the same management, namely nasogastric tube (NGT).

According to the literature, DRGF is observed in around 20% of patients following hepatic surgery [12], and it significantly affects the postoperative course and reduces postoperative quality of life [13, 14]. Therefore, it is crucial to gather more information about this common complication and investigate its predictive factors. The aim of the present study was to define the predictive factors for DRGF following PH.

All consecutive patients from 2010 to 2019 who underwent elective PH for a benign or malignant pathology at the Department of Visceral Surgery of Lausanne University Hospital (CHUV) were included. All patients after 2013 were managed according to an Enhanced Recovery After Surgery (ERAS) protocol for hepatic surgery [15]. Treatment decisions were discussed in a multidisciplinary meeting. Written informed consent was obtained for each patient.

Patients’ demographics, comorbidities, diagnosis, and previous surgeries or procedures (e.g., portal vein embolization, biliary stenting) were retrospectively collected. Additionally, perioperative data such as operation duration, epidural anesthesia, duration of vascular clamping, surgery extent, extrahepatic resections and associated radiofrequency were also collected. Postoperative morbidity and mortality, as well as length of stay (LOS), were assessed for each patient. Postoperative abdominal and non-abdominal complications, such as renal or pulmonary, were collected in our database. Complications were defined according to the Clavien-Dindo classification [16] of postoperative complications, and the Comprehensive Complication Index (CCI) [17] was calculated. Major complications were defined as Clavien-Dindo grade ≥IIIa.

The primary endpoint was DRGF, defined as the need for postoperative NGT insertion, consistent with a previous study [18]. To avoid diagnostic bias due to the clinical and therapeutical similarities that they share, PI and DGE were not separately investigated. Patients who had a NGT insertion during surgery and left the operation room with it, which is not a standard procedure in our center, were excluded. Blood loss was estimated by anesthesiologists based on the amount of liquid suctioned during surgery, after deducting liquid irrigations.

Liver failure was defined according to the International Study Group of Liver Surgery (ISGLS) definition (postoperative deterioration in the liver’s ability to perform its synthesis, excretion, and detoxification functions) [19]. This condition was practically defined by an increased international normalized ratio and concomitant hyperbilirubinemia observed on the fifth postoperative day. Biliary leak and post-hepatectomy hemorrhage were also defined according to the ISGLS [10, 20].

In our center, for the management of postoperative nausea and vomiting, we employ a multifaceted approach that aligns with international consensus recommendations [21]. During surgery, we utilize a 5-HT3 antagonist as the first-line agent, complemented by a low dose of dexamethasone as an additive to enhance antiemetic efficacy. Postoperatively, we continue to monitor patients closely, administering additional 5-HT3 antagonists and low-dose dexamethasone if needed.

We do not systematically administer medications to stimulate bowel function after liver surgery. In our center and protocol, early oral nutrition is considered a key component of postoperative recovery. On the day of surgery, patients are permitted to receive liquid food. Beginning on postoperative day 1, patients are encouraged to have a normal diet as long as they can clinically tolerate it.

Our institution employs a multimodal approach to pain management, which can include the use of intrathecal opiates. For open surgical procedures, we typically manage pain through a combination of epidural analgesia and first-step pain killers (paracetamol, metamizole). From postoperative day 3, at which point the epidural analgesia is often ceased, opioids are introduced if further pain relief is necessary. For minimally invasive procedures, the use of epidural analgesia is not routinely implemented. Instead, pain management is achieved through first-step [22] analgesics and additional opioids as required.

Continuous data were compared with the Student’s t test or Mann-Whitney U test as appropriate and were presented as mean (standard deviation, SD) or median (interquartile range, IQR). Categorical variables were analyzed with χ² test and presented as frequencies (%).

To assess the independent risk factors for DRGF, predefined variables were included in a multivariable binary logistic regression model. p value <0.050 was considered statistically significant. Analyses were performed with SPSS^® software, version 28.0.1.0 for Mac^® (IBM, Armonk, NY, USA).

A total of 501 patients were included. Patient demographics and comorbidities are listed in Table 1. Table 2 lists the diagnoses of the patients who underwent surgery. Most of the patients (54%; n = 270) were operated on for one or more liver metastasis, while 25% (n = 125) were operated on for a primary liver tumor (hepatocellular carcinoma or cholangiocarcinoma). DRGF occurred in 82 patients (16%). There were no significant differences between the two groups regarding preoperative data except that patients with DRGF had previously undergone liver surgery more frequently than those without DRGF (8/82 = 10% vs. 8/419 = 2%; p = 0.002). It should also be noted that a history of gastrectomy or colectomy was not more frequent in patients with DRGF.

Out of the total patients, 12% (n = 60) suffered from a biliary leak, including 7% (n = 36) with leaks on the hepatic section surface and 5% (n = 24) from the biliary anastomosis. Additionally, 3% (n = 13) experienced hemorrhage, while 2% (n = 9) developed hepatic failure. Out of the patients who underwent a major resection, 64% (153/238) experienced complications, while out of those who underwent a minor resection, 37% (97/263) experienced complications.

In our cohort, 22.4% underwent a minimally invasive approach and it was observed that the DRGF incidence was significantly lower in this group compared to open procedures (5/112 = 4.5% vs. 77/389 = 19.8%, p < 0.001). Among DRGF patients, 17% (n = 14) needed a second NGT placement. Median time of NGT was 2 days (IQR 1–3) and 5% of the patients (n = 23) received gastrografin to stimulate the return of gastrointestinal function. There were no significant differences in demographic and intraoperative data between patients who had the NGT placed once or twice. DRGF incidences were similar before and after Enhanced Recovery after Surgery implementation (16/78 = 21% vs. 66/423 = 16%, p = 0.281). No difference in DRGF incidence was found in the case of right or left PH (33/145 = 23% vs. 26/125 = 21%, p = 0.768).

Intraoperative and postoperative characteristics are listed in Table 3. Major PH (≥3 segments) was performed in 238 patients (48%) and it was significantly associated with DRGF (55/238 = 23% vs. 27/263 = 10%, p < 0.001). DRGF occurred more frequently in patients with preoperative embolization (26/88 = 30% vs. 55/407 = 14%, p < 0.001), biliary anastomosis (20/48 = 42% vs. 61/450 = 14%, p < 0.001), and extrahepatic resection (37/108 = 34% vs. 45/393 = 11%, p < 0.001). Extrahepatic resection included 14 colorectal resections (2.8%), 3 gastric resections (0.6%), and 91 others. All the extrahepatic resections performed are listed in Table 4. None of the patients who underwent a colonic resection experienced an anastomotic leak. Patients with DRGF had longer median operation duration (374 vs. 263 min, p < 0.001), more leaks on the hepatic section surface (16/82 = 20% vs. 20/419 = 5%, p < 0.001), and a higher incidence of leaks from the biliary anastomosis (11/82 = 13% vs. 13/419 = 3%, p < 0.001), and a higher rate of extra-abdominal complications (33/82 = 40% vs. 57/419 = 14%; p < 0.001). Additionally, they had a higher median blood loss (1,088 vs. 701 mL, p < 0.001). DRGF patients developed more pneumonia (14/22 = 64% vs. 8/22 = 36%, p < 0.001) and had longer median LOS (19 vs. 8 days, p < 0.001).

On multivariable analysis, operation duration (OR 1.005, 95% CI 1.002–1.008, p < 0.001), major PH (OR 3.606, 95% CI 1.931–6.732), and postoperative biloma/biliary leak (OR 6.419, 95% CI 3.019–13.648, p < 0.001) were independently associated with DRGF occurrence (Table 5).

The findings of the present study that investigated the predictive factors of DRGF after PH showed that duration of surgery, major PH and occurrence of biloma or biliary leak are three independent predictors of DRGF. Moreover, patients who experienced DRGF developed more pneumonia and had a longer LOS.

In the present cohort, DRGF occurred in 16% of the patients. This includes the incidence of postoperative DGE and PI, as our definition was based on the insertion of a NGT. A previous study [12] showed an incidence of 20% of DRGF after pH. As in our study, the authors defined DRGF as the postoperative need to insert a NGT. DGE and PI have been studied separately for other types of abdominal surgery. In general, PI was found in 10–30% of patients who underwent abdominal surgery [18, 23, 24], especially in colorectal surgery. DGE has been studied mainly after surgery of the upper GI tract as a consequence of vagal nerve injury, with an incidence found in the literature of 5–25% after gastric surgery [25‒27], 20–40% after pancreatic resection [28‒30] and 1–2% after living donor hepatectomy, especially if left-sided [31, 32]. Indeed, studies have shown that left hepatic resections are more associated with DGE than right resections [31, 33‒35]. The underlying mechanisms of gastric stasis after left PH can be explained by many factors. First, during a left-sided PH, the surgical procedure involves manipulation of the left lobe of the liver, which is closer to the stomach. This can disrupt the normal functioning of gastric and vagal nerves, leading to DGE [36, 37]. Moreover, left-sided PH causes an inflammatory response and the development of adhesions in the abdominal cavity. These postoperative changes can interfere with gastric motility, contributing to DGE [31, 33, 34]. In our study, there was no significant difference between right and left resection, and this is probably due to the fact that we included both DGE and PI in the definition of DRGF.

Age has been associated with increased morbidity and mortality after liver surgery [3, 38, 39]. The implementation of ERAS has significantly reduced this risk [15], but it is not clear whether it has also reduced the risk of DRGF specifically in the elderly population. In our study age was not associated with the development of DRGF. On the other hand, Arfa et al. [12] found a significantly higher incidence of DRGF in this patient group but further studies are needed to investigate the specificities of this population.

In our study, DRGF occurred more frequently in patients who underwent preoperative embolization, biliary anastomosis, and extrahepatic resection. Previous studies have shown that patients who had an embolization to surgery underwent a more complex resection, had a longer operative time and also had a higher rate of intraoperative transfusion [40, 41]. This suggests that patients undergoing preoperative embolization have more advanced disease that leads to a more difficult surgery and an increased risk of complications. Similarly, patients who undergo biliary anastomosis often have major liver resections and therefore longer and more complicated surgeries [42, 43].

Patients who have had a biliary anastomosis are also at risk of biliary leakage, with an incidence between 6 and 30% depending on the series [44‒47]. The literature shows that these patients are at greater risk of complications, particularly bleeding and liver failure [47].

Patients with DRGF had longer median operation duration and higher median blood loss. These findings are also found in the literature. Indeed, Arfa et al. [12] showed a significantly higher operative duration and blood loss for patients who experienced DRGF, but these two factors were not independently associated with DRGF. In the present study, operation time was independently associated with DRGF. Nie et al. [48] also showed that blood loss during surgery and operating time were two risk factors for developing DGE in abdominal non-gastroduodenal surgeries. These risk factors are also found for postoperative PI after abdominal surgeries [24, 49].

The current study has several limitations. First, our definition of DRGF included DGE and PI for practical reasons, even though these conditions have distinct pathophysiological mechanisms. It would be valuable to conduct separate studies focusing on each of these pathologies, potentially including only patients who received a radiological diagnosis. However, implementing such an approach would necessitate the exclusion of a substantial portion of our patient population. Second, DRGF was defined as the need for NGT, which means that patients who experienced symptoms of DRGF (nausea, vomiting, constipation) but did not necessitate the insertion of a NGT for treatment were not considered. Finally, only a minority of patients benefited from a minimally invasive laparoscopic approach, and most of these patients underwent minor liver surgery. This may therefore have influenced the results by minimizing the benefits of a laparoscopic approach.

This study provides a precise understanding of DRGF incidence, which is valuable for obtaining patient consent preoperatively and for setting appropriate postoperative expectations. Second, it prompts a more thorough examination of pneumonia in DRGF, even if it is expected intuitively, ensuring comprehensive postoperative care. Finally, it helps identify independent predictors of the condition, facilitating targeted interventions and personalized treatment plans.

In conclusion, DRGF was frequent in the present cohort (16%) and was associated with longer length of stay. Surgery duration, major PH, and biliary leak were found as independent predictors of DRGF.

This study protocol was reviewed and approved by the Cantonal Commission on Ethics in Human Research (CER-VD), Approval No. 2017–01169. Patient written informed consent was obtained.

The authors have no conflicts of interest to declare.

The authors declare that there are no funding sources or financial support for this study.

G.P. collected the data, performed the analysis, and wrote the manuscript. G.-R.J. verified the analytical methods, supervised the data analysis, and supervised the project. I.L., E.U., E.M., and N.H. helped supervise the project. All authors discussed the results; provided critical feedback; and helped shape the research, analysis, and manuscript.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author (G-R.J.).