Introduction

Liver cancers are highly malignant, often fatal, and have a global occurrence [1, 2]. They are classified, based on the source of tumor cells, into either hepatocellular carcinomas or cholangiocarcinoma, with the latter having more aggressiveness and poorer prognoses [2,3,4]. Anatomically, cholangiocarcinoma is divided into extrahepatic cholangiocarcinoma, hilar cholangiocarcinoma, and the more common intrahepatic cholangiocarcinoma (ICC). Most incidences of ICC are found in individuals aged 60 and older because the typical age for the onset of ICC is between 50 and 70 years [5]. Despite patients aged 60 and older having a low tolerance for intensive surgery, inferior cardiopulmonary function, and high risk of postoperative complications, complete surgical excision remains currently preferred treatment for all ICC patients [6, 7]. This surgical excision is performed via either conventional open hepatectomy (OH) or laparoscopic hepatectomy (LH). With the advance of modern medicine in terms of minimally invasive surgery and laparoscopic techniques in particular, LH has several advantages such as mild trauma, rapid recovery, lower intensity of pain, and short hospital stays [8, 9]. Nonetheless, LH may nonetheless lead to increased risk of cardiorespiratory and unpredictable problems [10]. Currently, the benefits of LH for ICC patients who were older remain poorly understood. Therefore, the present study examined short- and long-term outcomes of patients aged 60 and older with ICC who had undergone either OH or LH. The study also assessed potential risk factors for prognosis.

Materials and methods

Patient enrollment

Between January 2011 and October 2021, 107 ICC patients who had undergone hepatectomies at Qilu Hospital of Shandong University, Jinan, China, were enrolled, having fulfilled the following inclusion criteria: (1) patient with ICC confirmed by postoperative pathology, (2) patient underwent complete surgical tumor resection without macroscopic residual tumor (R0 or R1) and absence of distant metastasis, and (3) patient was monitored until date of mortality/last check-up. Surgical procedures were selected based on the following criteria: (1) According to whether it is minimally invasive surgery or not, the patients and their families chose the open or laparoscopic approach independently, (2) the surgeon evaluated the feasibility of laparoscopic surgery based on the location of the tumor, and (3) some patients who chose open surgery because of the difficulty of surgery were excluded (two patients who converted to open surgery due to intraoperative massive bleeding). In total, 29 patients who only underwent laparoscopic surgery during tumor resection were divided into the LH group, and 78 patients who underwent laparoscopic exploration or open surgery were divided into OH group. The present research represents a retrospective study, and the data used in this study were queried from Qilu Hospital of Shandong University databases, following approval by the hospital.

Preoperative preparation

Both liver function tests and preoperative hematology examinations were conducted on all patients. Further, patients were regularly provided with support therapy and monitored using contrast-enhanced computed tomography (CT), magnetic resonance imaging (MRI), electrocardiograms (ECG), echocardiography, and specifically monitored via automatic 3D liver organ reconstruction that was used for on-demand volumetric analysis. Before surgery, patients fasted immediately prior to gastric tube insertion. Lastly, preoperative assessments and re-operative health education were required of all patients.

Surgical technique

All patients were placed in the supine position after general anesthesia was administered. Routine surgical disinfection and draping were performed. For the LH group, pneumoperitoneum was established using the infraumbilical port, followed by insertion of three or four additional trocars. Before liver resection, tourniquets were prepared to possibly block the portal vein. Sterile gauze was used to expose the operative field and clean the surgical area. Standard laparoscopic instruments and an ultrasonic scalpel were used for surgical procedures such as partial hepatectomy, segment hepatectomy, and lobectomy whose election was determined by tumor location. If necessary, intraoperative ultrasonic and indocyanine green were used to locate tumors. For the major hepatectomy which involved 3 liver segments, and extended liver resection which involved more than 3 liver segments, such as left hemihepatectomy or right hemihepatectomy, we preferred the extrahepatic Glissonean approach. Minor liver resection was defined as the removal of one or two segments, the hepatic parenchymal transection approach was the preferred choice. For the OH group, traditional open hepatectomy was done with an anti-L-shaped incision of ca. 30cm. An ultrasonic scalpel (similar to that used in the LH group) was used to cut liver tissue, whereas absorbable and home-lock clamps were used to occlude the bile duct and vascular system. Bipolar electro-coagulation was regularly performed for hemostasis. Lymph node dissection (LND) is generally not performed in all patients at the center of our study, conversely, only performed in patients with enlarged lymph nodes on preoperative imaging or detected during intraoperative examination.

Postoperative management

After surgery, all patients received professional postoperative management, nutritional support, rehydration, symptomatic support, and standard postoperative analgesic plans. On the first and third postoperative days, blood routine, coagulation tests, liver, and kidney functions and other indicators were examined. Additionally on the first postoperative days, gastric tubes were removed and on the fourth postoperative days, abdominal CT scans were done to assess patients’ intra-abdominal condition.

Data collection and definitions

Patient’s medical records, which included demographic characteristics, comorbidities, preoperative and postoperative hematology and chemistries, intraoperative data, postoperative, and follow-up data, were collected retrospectively. Postoperative follow-ups were performed once every 3 months via telephone interviews. Overall survival (OS) was defined as the duration between surgery and death, and disease-free survival (DFS) was defined as the post-surgery duration that had no signs of disease. The 8th edition of AJCC/UICC TNM staging system and the Clavien-Dindo complication classification system were used to evaluate postoperative complications. According to the WHO latest principles of antibiotic classification, the antibiotic usage was divided into two categories: nonrestricted antibiotics and other antibiotic (restricted antibiotics and special class antibiotic).

Statistical analysis

Categorical variables were presented as percentages and the statistical significance of their comparisons was assessed using either χ2 or Fisher’s exact tests. Continuous variables were expressed as medians/25th–75th percentiles/means/± standard deviations, and the statistical significance of their comparisons was assessed using either Student’s t or Mann–Whitney U tests. OS and DFS curves were plotted using Kaplan-Meier (KM) curves. Risk factors for OS and DFS were analyzed by Cox regression. All statistical analyses were done in SPSS Statistics V.25 (IBM SPSS Software) and/or R V.3.5.3, and a 0.05 alpha level was used for all tests.

Results

Patient characteristics

The baseline characteristics of patients in the two groups are summarized in the Table 1. Patients in LH and OH groups did not significantly differ (p>0.05) in terms of sex, age, body mass index (BMI), hypertension, diabetes, smoking, alcohol consumption, and hepatitis B virus infection. In addition, there were 11 (14.1%) and 5 (17.2%) patients with liver cirrhosis in the OH and the LH group, respectively, which was no significant differences (p=0.921). Moreover, no significant differences (p>0.05) in portal hypertension, ascites, preoperative antitumor treatment, and all preoperative blood tests except preoperative albumin (Alb) levels, which were significantly lower (p=0.039) in the OH compared to the LH group. However, the mean values of Alb for both groups were over the lower limit of normal, which is 40 g/L in our research center (Table 1).

Table 1 Comparison of the baseline characteristics between OH and LH groups
Full size table

Pathological findings

A comparison of pathological findings between the OH and LH groups was as summarized in Table 2. Although the proportion of patients with lymph node metastasis was higher in the OH than the LH group (20.5 vs. 6.9%), this difference was not statistically significant (p=0.167). Other parameters without statistically significant differences in the two groups were pathological differentiation, tumor number, tumor diameter, TNM stage, surgical margin, microscopic microvascular invasion, microscopic perineural invasion, and satellite foci. It should be mentioned that no differences in major hepatectomy ratio were found between the two groups of our study (64.1 vs. 58.6%, p=0.167).

Table 2 Comparison of the pathologic findings between the OH and LH groups
Full size table

Surgical outcomes

Surgical outcomes for both OH and LH groups are summarized in Table 3. Incision lengths were obviously significantly shorter in the LH than OH group (5.7±2.3 vs. 20.1±1.5 p=0.000). Differences in operation time, albumin infusion, and intraoperative and postoperative transfusions were not statistically significant (p=0.907, p=0.070, p=0.185, and p=0.555, respectively). Similarly, R0 resection (92.3 vs. 93.1% p=1.000), liver door block (34.6 vs. 55.2% p=0.054) rates, and LND (47.4 vs. 27.6% p=0.064) rates were not significantly different between the two groups. In addition, there was no significant difference in the comparison of the number of lymph nodes between the groups (p=0.165). The LH group, had no intraperitoneal infections, incision-related complications, delayed gastric emptying, and bile leakage nor ICU admissions. The number of other postoperative complications—arrhythmology (0 vs. 3.4% p=0.271), peritoneal effusion (12.8 vs. 6.9% p=0.604), lung infection (14.1 vs. 3.4% p=0.227), myocardial infarction (0 vs. 6.9% p=0.072), and heart failure (2.6 vs. 10.3% p=0.122)—did not significantly differ between the two. Furthermore, grading of antibiotic was not different in both groups (p=1.000). Table 3 summarizes results of hematological examinations undertaken on the first postoperative day—total bilirubin levels (TBIL) were significantly higher in the OH than LH group (p=0.01). Plasma ALT and AST levels (median 297.0 and 261.5, respectively) in the OH group were higher than those (median 208.0 and 242, respectively) in the LH group (p=0.271 and p=0.446, respectively).

Table 3 Comparison of the surgical outcomes and follow-up outcomes between OH and LH groups
Full size table

It was found that postoperative hospital stays were significantly shorter in the LH than OH group (8.6 days vs. 11.6 days, p=0.009). However, hospital costs did not significantly differ between the two (7.7W RMB vs. 7.2W RMB, p=0.502). For both groups, no deaths were recorded within the first 30 days post-surgery.

Long-term outcomes

Comparisons of long-term outcomes between the OH and LH groups are shown in Table 3 and Figs. 1 and 2. It was found that a higher proportion of patients relapsed (59.0 %) and died (46.2 %) in the OH group as compared with that in the LH group (37.9 % and 31.5 %, respectively). However, it was evident that the differences of relapse and death in the two groups were not statistically significant (p > 0.05). In patients who relapsed, the hepatic and extrahepatic recurrence (52.2% and 54.5%) was the main site of relapse in both groups, followed by hepatic recurrence (32.6% and 36.4%). The total number of patients who received treatment following recurrence are 41 and 9 patients in two groups, respectively (p > 0.05). Among the patients with recurrence in the two group, 4 patients (8.7%) and 2 patients (18.2%) received radiofrequency ablation (RFA), 11 patients (23.9%) and 2 patients (18.2%) received repeated liver resection, 19 patients (41.3%) and 3 patients (27.3%) received oncologic palliative therapy, 3 patients (6.5%) and 0 patients (0.0%) received repeated liver resection and oncologic palliative therapy, and 4 patients (8.7%) and 2 patients (18.2%) received RFA and oncologic palliative therapy.

Fig. 1
figure 1

Comparison of overall survival between the two groups

Full size image
Fig. 2
figure 2

Comparison of disease-free survival between the two groups

Full size image

The mean follow-up time was similar between the OH group and LH group (17.2±16.4 vs. 13.3±12.3, p = 0.254). The median OS and DFS for the OH group were 20.5 and 13.0 months, respectively, whereas for the LH group was 27.1 and 18.9 months, respectively (Figs. 1 and 2). These differences between the two groups were also not statistically significant (p = 0.53 and p = 0.35, respectively).

Univariate and multivariate analyses of factors associated with OS and DFS

Results of univariate and multivariate analyses of variables for OS and DFS are summarized in Tables 4 and 5, respectively. CA-19-9, BMI, neutrophil count, TNM stage, platelet count, AST, pathological differentiation, lymphatic metastasis, satellite foci, and liver capsule invasion were independent prognostic factors for OS (p<0.05), and CA-19-9 (HR 1.002, 95% CI 1.001–1.003, p=0.000) and moderate differentiation (HR 0.344, 95% CI 0.159–0.744 p=0.007) were independent risk factors for OS. CA-19-9, TNM stage, neutrophil count, pathological differentiation, lymphatic metastasis, and liver capsule invasion were independent prognostic factors for DFS (p<0.05), and CA-19-9 (HR 1.001, 95% CI 1.000–1.002, p=0.006) and neutrophil count (HR 1.326, 95% CI 1.128–1.558, p=0.001) were independent risk factors for DFS.

Table 4 Univariate and multivariate analyses of factors associated with overall survival rates
Full size table
Table 5 Univariate and multivariate analyses of factors associated with disease-free survival rates
Full size table

Discussion

ICC is the second most common primary malignant hepatic tumor [11] and most remain undiagnosed at early stages due to lack of typical distinct symptoms and have poor prognoses [3, 6]. The same phenomenon was observed in our study, where patients with relapse had a high mortality rate in both groups (36/46 and 9/11, respectively). Moreover, these tumors remain most neglected in the patients aged 60 and older, due to low awareness of the illness.

Surgical resection, especially R0, is the main treatment for ICC patients despite its considerable risk of recurrence [3, 6, 7]. In this study, most patients in both groups had R0 resection, indicating that laparoscopic surgery was now relatively feasible for ICC treatment. Rather than only in our research, many research centers had demonstrated the same attitude [12,13,14,15]. Relatedly, LH, when compared to OH for ICC treatment, had a higher R0 resection, more favorable short-term outcomes and lower LND rates [8, 9, 14, 15]. Furthermore, it was found that the LND rates in this study were lower in the LH as compared with the rates in OH group (Table 3), but the differences were not statistically significant.

Lymph node metastasis (LNM) had been established as a critical risk factor for prognosis in ICC [16], and the controversy of lymph node dissection for ICC had been attracting great attention in recent years. Advocates of routine LND in ICC patients suggest that LND not only prolongs OS and DFS but also allows for accurate lymph node staging, which can help in determining the patient’s prognosis and developing subsequent adjuvant treatment plans [17, 18]. While opponents argue that LND fails to improve OS or DFS in such patients and was responsible for the increased risk of postoperative complications [19]. Therefore, we performed the cox regression analysis to explore the clinical significance of LND in our experiments. The results revealed that both LND and the number of LND may not be an independent prognostic factor for OS and DFS. This may suggest that LND does not affect the prognosis and outcome in patients aged 60 and older with ICC.

Mean tumor diameters were larger in the OH than LH group, but this difference was not statistically significant. This difference may be because, logically and ethically, conventional open surgery was recommended for patients with large tumors. However, the rate of major hepatectomy via laparoscopic in LH group is comparable to OH group in our research center, which benefitted greatly from advances in laparoscopic instruments and techniques and indicated that laparoscopic major hepatectomy is also can performed. The significant differences in incision length and the duration of postoperative stays between the two groups (Table 3) corroborated previous studies [8, 12]. An advantage of shorter incision lengths in LH when compared to OH is prevention of possible complications and pain reduction. Indeed, the LH group had a zero incidence of incision-related complications. Postoperative infection, particularly lung infection in elderly patients, is an increasingly common clinical complication [20,21,22]. It is evident that no significant distinction was indicated in antibiotic usage. In this case, the rates of lung and intraperitoneal infections were lower in the LH than OH group. Therefore, a worthwhile research study would be to explore how laparoscopic surgery can be used to reduce postoperative infections in the elderly.

Owing to the dismal prognosis of ICC, it is urgent to find prognostic indicators to improve the prognosis. Perioperative level of Alb is reported as a possible independent prognostic factor for ICC [23,24,25]. In our present study, the preoperative levels of Alb significantly differed between the two groups (Table 1). However, we tentatively suggest that Alb played little role for the long-term outcomes in the study. There are two explanations. First, mean Alb levels for both groups not only slightly differed but also were over the lower limit of normal (Table 1). Second, a substantial proportion of patients lacked sufficient c-reactive protein (CRP) levels; thus, CRP/Alb ratios could not be calculated [24,25,26]. The total bilirubin level (TBIL) is a significant index of liver function [27,28,29,30], and thus, TBIL was monitored and high TBIL was promptly treated. Additionally, TBIL was an important independent prognosis predictor for ICC patients [31,32,33] and was significantly higher in the OH than LH group on the first postoperative blood tests (Table 3). However, TBIL did not differ between the two before operations, further demonstrating laparoscopic in lieu of open surgery as a means to prevent high TBIL. Body mass index (BMI), which is calculated from weight and height, is highly correlated with obesity. Obesity is a major concern that is correlated with increasing risks of cancers, including ICC [34, 35]. Likewise, the BMI is also known to be associated with the prognosis of ICC [36]. In the present study, we demonstrated that BMI was an independent prognostic factor for ICC (Tables 4). The ratio of the neutrophil count and other inflammatory biomarkers can predicate the prognosis of an ICC patient [37, 38]. Neutrophil count also affect prognoses and we also demonstrated this (Tables 4 and 5). The understanding of those risk factors and protective factors associated with prognosis is important to guide future interventions directed at improving long-term prognosis for of patients with ICC.

Research on treatments that ensure long-term survival of patients is high relevant for clinical cancer. Today, surgery-based comprehensive therapy combined with chemotherapy and molecular targeted therapy is the recommended basic route. In terms of surgery-based comprehensive therapy, this study demonstrated that ICC patients who underwent LH achieved long-term outcomes comparable to those of patients who underwent OH, and this corroborated previous studies [8, 12, 13]. Taken together, LH was a safe and feasible alternative treatment for ICC patients aged 60 and older, which achieved short and long-term outcomes that were comparable and even superior to those of OH.

There are two limitations in the study should be point out. Firstly, it was a retrospective analysis. Second, it had a small sample size (n=107) due to insufficient medical records for a large proportion of patients who were from other research centers. The latter especially limited the study as the robustness of the employed methods increases with the amount of systematic information and sample size.

Conclusion

In conclusion, LH was safe, reliable, and feasible for the treatment of ICC patients aged 60 and older. We also found that the rate of major hepatectomy and LND in the LH group can achieve the same effects as the OH group. Furthermore, patients in the LH group had better short-term clinical outcomes as compared with that in OH whereas the long-term prognoses of patients in the two group were statistically comparable to each other.