Surgery for hepatocellular carcinoma and intrahepatic cholangiocarcinoma: milestone changes in the last two decades potentially affecting current guidelines

INTRODUCTION

Until the ‘90s, the only possible technique for performing liver resections was laparotomy. In 1991, the first laparoscopic liver resection was performed, thus opening up a new promising field of liver surgery^[1]. During the next thirty years, minimally invasive liver resection (MILR) spread worldwide due to the benefits seen in patient outcomes. MILR has been reported beneficial in terms of blood loss, transfusion rates, pain control, complications, and length of hospital stay^[2-7].

Gradually, MILR has become feasible, safe and efficient for the treatment of both benign and malignant liver diseases, including those requiring complex hepatectomies^[8-10]. The diffusion has been slow but steady, and thus MILR has become a well-established field of hepatobiliary surgery and is considered a fundamental change in recent years^[11].

The purpose of this narrative review is to describe the relevance of MILR for the treatment of the most common primary liver tumors, namely hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC). The treatment for the two types of tumors is different because they both have specific oncological and surgical characteristics that must be considered in the application of minimally invasive techniques for their resective treatment.

REVIEW STRATEGY

The methodology implied a literature search performed in the PubMed database conducted with the search words (“HCC”, or “hepatocellular carcinoma”, or “cholangiocarcinoma”, or “intrahepatic cholangiocarcinoma”) AND (“laparoscopic”, or “minimally invasive”, or “robotic”, or “robot-assisted”). The definition of “minimally invasive” included pure laparoscopic and robotic techniques, which are the most common minimally invasive approaches for liver surgery; hand-assisted, hybrid and laparoendoscopic single-site techniques were not considered. Laparoscopic and robotic surgery have many similarities: both use small incisions, a camera that magnifies the view, and require the surgeon to operate instruments from outside the patient’s body. However, laparoscopic devices are manipulated directly by hands, whereas robotic instruments use a console as part of an advanced surgical system with arms moving around the surgical field.

A language restriction was applied to include English literature only and publications up to September 2022 were considered. The authors have selected and included the most relevant literature evidence to support the illustrated hot topic concepts. It should be emphasized that the evidence on MILR for HCC and ICC is essentially supported by retrospective studies. This is in line with the general paucity of prospective studies in the surgical field. While the rigorous form of RCT is the ideal method for providing evidence, prospective surgical trials are associated with several difficult obstacles to overcome. Timing and planning are complex as surgical techniques evolve over time; perioperative care and surveillance often differ between centers; surgical randomized controlled trials are not usually carried on in stages (as drugs trials are), and selecting the most appropriate endpoints can therefore be difficult; enrollment can be challenging because patients are often unenthusiastic about invasive procedures unless the surgeon has unequivocal recommendations.

HEPATOCELLULAR CARCINOMA

The indications for the treatment of HCC are well defined by the Barcelona Clinic Liver Cancer algorithm, which has been updated over time^[12]. Despite the significant evolution, the eligibility for liver resection remains limited to very early and early HCC in compensated cirrhosis mainly due to two factors: tumor burden (advanced, multifocal disease) and functional liver reserve (decompensated cirrhosis, portal hypertension). Wrong indications can lead to unsafe perioperative outcomes and questionable oncological benefits. MILR is among recent advances that play a role in increasing the safety and efficacy of liver surgery for HCC and potentially expanding the pool of patients eligible for liver resection.

Research into the specific role of MILR for HCC began twenty years ago with some early studies based on a few patients from pioneering surgical centers. As for other liver diagnoses, the general benefits of MILR compared to laparotomy have gradually emerged even for HCC, such as the reduction of blood loss and the shorter length of hospital stay. More importantly, MILR in this oncological setting appeared to be associated with specific advantages in terms of reduced incidence of postoperative hepatic decompensation and ascites^[13,14]. Indeed, early in 2010, Aldrighetti et al. showed lower blood loss and shorter length of hospitalization compared to open (258 mL vs. 617 mL, P = 0.008; 6.3 vs. 9 days, P = 0.039). In the same year, Tranchart et al. reported decreased ascites following MILR than laparotomic resection (7.1% vs. 26.1%, P = 0.030). Evidence has accumulated over time with many publications based on increasing numbers of HCC treated, and to date, several meta-analyses have pooled data from retrospective comparative studies on minimally invasive versus open liver resections^[15-22] [Table 1].

In most of these publications, laparoscopy was confirmed to be associated with reduced blood loss and transfusion rates, reduced overall morbidity and length of stay, with non-inferior perioperative outcomes compared to open surgery. Importantly, a meta-analysis looking at 1618 patients with HCC showed a 20% reduction in the hazard ratio of death for patients with cirrhosis who underwent MILR^[23]. Furthermore, several meta-analyses have demonstrated the advantage of MILR with regard to specific complications of liver surgery for HCC (reduced incidence of postoperative hepatic decompensation and ascites). For example, in a systematic review of high-quality case-matched studies published in 2018, Chen et al. reported that ascites was less in MILR than in open resections, regardless of whether they underwent minor or major hepatectomies (RR = 0.42, P < 0.010) and patients undergoing laparoscopy were less likely to suffer from liver failure (RR = 0.41, P < 0.010)^[24]. The most frequently reported reasons for the reduction of hepatic decompensation in the MILR series are the avoidance of interruption of abdominal collateral and lymphatics vessels by small incisions and limited mobilization of the liver, thus allowing the preservation of portosystemic wall shunts and thereby limiting postoperative portal hypertension, subsequent ascitic decompensation, and impaired liver function. Moreover, minimally invasiveness is able to limit postoperative ascites by reducing the laparotomy-related fluid dispersion, thus facilitating intra- and postoperative fluid and electrolyte balance^[25]. In addition, the reduced surgical stress response observed for many minimally invasive surgeries may also contribute to reducing the biological impact of surgery and underlie better perioperative outcomes than open surgery^[26]. As a further explanation of this advantage, to date, there is no evidence that MILR allows for being more selective in parenchymal resection. However, considering advances in laparoscopic anatomical resections, the increasing use of intraoperative indocyanine green fluorescence imaging and the diffusion of robotics, these techniques may allow further refinements to reduce the amount of nontumor tissue resected and further enhance the benefits on postoperative ascites and liver decompensation.

It is worth mentioning the result of a publication that aimed to analyze the outcome of MILR in patients with early and intermediate stage disease according to BCLC and to define the differential benefit with open. The differential benefit was evaluated taking into account blood loss and morbidity rate as outcome indicators and resulted in a greater advantage of MILR in intermediate stage patients compared to early stage patients for both the parameters (-270 mL vs. -150 mL, pure differential benefit: 120 mL, P = 0.028; -4.6% vs. -7.4%, differential benefit: 2.8%, P = 0.042). The conclusion was that the favorable biological scenario associated with MILR allows for greater benefits in the context of more advanced liver disease, which could potentially promote stage migration in suitable patients^[27].

Most studies and meta-analyses evaluating the oncological efficacy of MILR for HCC have documented short- and long-term outcomes comparable to open, even for major hepatectomies^[28-30]. Surrogate oncological outcomes such as radical resection rates and resection margin width are reported with no inferior outcomes due to the adoption of minimally invasive approaches, as well as recurrence-free, disease-free and overall survival rates are reported as similar. Of note, some recent publications have reported even better oncological outcomes for MILR than for open. Lower tumor recurrence rates for MILR (OR: 0.65, P = 0.008) and better 1-y disease-free (HR: 0.71, P = 0.02) and 5-year overall survival (HR: 0.60, P = 0.004) were shown by a meta-analysis of more than 800 HCC patients published in 2018 by Goh et al., who linked these findings to reduced blood loss and lower rates of positive resection margins^[31]. In 2021, another study compared survival outcomes of MILR and open resections in the treatment of HCC by reconstructing time-to-event data for meta-analysis with Kaplan-Meier curves of high-quality propensity-score matched studies. The results suggested that MILR may improve RFS in HCC patients undergoing minor hepatectomy (HR: 1.44, P < 0.001)^[32]. These data support the possibility that MILR may have a favorable impact on the oncological prognosis of resected patients by reducing tumor recurrences and consequently improving disease-free and overall survival.

HCC patients with advanced cirrhosis and portal hypertension are those in whom open liver resections carry a substantial risk of mortality and relevant complications. MILR is increasingly being adopted to expand liver resection to these categories of HCC patients. In 2018 a multicenter study showed no significant differences in short-term surgical outcomes between Child A and Child B patients undergoing MILR: blood loss (400 mL vs. 350 mL, P = 0.856), transfusion rate (15% vs. 12%, P > 0.05), postoperative mortality and morbidity rates (2% vs. 4%, P = 0.491; 39% vs. 36%, P = 0.467) including liver-specific complications as ascites and liver decompensation were all similar (11% vs. 12%, P = 0.562; 3% vs. 4%, P = 0.595)^[33]. In the same study, clinically significant portal hypertension was not among the independent risk factors for major morbidity (or: 1.45, P = 0.287). The advantages of MILR in advanced Child-B cirrhotic patients have been confirmed by an international multicenter propensity score-matched study reporting decreased blood loss (110 mL vs. 400 mL, P = 0.004) and complications including major morbidity (38% vs. 51%, P = 0.041; 7% vs. 21%, P = 0.010) compared to open^[34]. The long-term impact of MILR for HCC patients with clinically significant portal hypertension has been evaluated in some studies showing 1-y and 3-y overall survival similar to patients without portal hypertension (100% and 87%, respectively, P = 0.800)^[35,36]. In light of the above findings, MILR can be considered a means with a protective effect against major postoperative complications, ascites and postoperative liver failure even in HCC patients with advanced cirrhosis and portal hypertension. As such, its role in expanding indications for resection is being considered.

The results of minimally invasive techniques for liver resections have been well summarized by various expert consensus. In 2017, the Southampton consensus led to the production of clinical practice guidelines for MILR, including its application in the context of HCC. It was pointed out that MILR for HCC is associated with reduced blood loss, transfusion rate, postoperative ascites, and liver failure and hospital stay with comparable to open operation time, disease-free margin, and recurrence rates and is an independent factor in reducing the complication rates, and that laparoscopic approach appears to reduce the incidence of postoperative ascites, liver failure and morbidity assessed in terms of "Comprehensive Complication Index”. In that document, the use of minimally invasive approaches for the surgical treatment of HCC was endorsed by experts in the field based on literature evidence and clinical experience^[37].

In summary, MILR for HCC has been performed largely worldwide, and the results have been investigated with propensity score matching studies and meta-analyses that included a relatively large number of patients. Compared to open resections, MILR shows advantages in perioperative outcomes and at least comparable oncologic outcomes. Although the majority of studies involve patients with compensated cirrhosis, some reports have demonstrated the feasibility of MILR in selected Child B/C patients. Further studies should refine the selection of patients with advanced cirrhosis who benefit most from a minimally invasive approach.

INTRAHEPATIC CHOLANGIOCARCINOMA

Cholangiocarcinoma is the second most common type of primary liver cancer after hepatocellular carcinoma. Among treatments of ICC, surgery remains the only potentially curative solution among existing therapeutic modalities^[38-44].

Despite continuous growth, the experience of MILR for ICC is still limited, and its adoption is not well established yet, mainly because of the technical and oncological hurdles associated with the frequent need for major hepatectomy and the indication of adequate lymphadenectomy. In the Southampton Consensus Guidelines, the role of MILR for the management of ICC has not been addressed, and - at present - it is considered in the initial phase of exploration. However, evidence is progressively accumulating and leading to the assumption that - in the near future - MILR will acquire a well-defined role in the treatment of ICC as MILR for other diseases.

Relatively small single-center series have reported the feasibility of MILR with lymphadenectomy for ICC, showing non-inferior results compared with open, with some advantages. In 2015, Uy et al. reported lower blood loss for 11 MILR for ICC (325 mL vs. 750 mL, P = 0.024), with similar morbidity (9.1% vs. 26.9%, P = 0.228) and width of resection margin (0.6 cm vs. 0.2 cm, P = 0.125)^[45]. Likewise, in 2016, Lee et al. published on 14 MILR for ICC, showing lower blood loss compared to open (325 mL vs. 625 mL, P = 0.006), and no differences in morbidity (21.4% vs. 19%, P = 1.000) and mortality (0 vs. 0, P = 1.000). As for oncological outcomes, both overall and recurrence-free survival were similar at 3 years with a median follow-up of 21 months (75.7% vs. 84.6%, P = 0.672; 56.7% vs. 76.9%, P = 0.456)^[46]. In the same year, Ratti et al. analyzed 20 MILR for ICC matched with 60 open. In this series, 85% of patients underwent major hepatectomy. There was no difference in operative time (210 min vs. 180 min, P = NS), complications (15% vs. 13.3%, P = NS), mortality (0 vs. 0, P = 1.000), and radical resection rates (100% vs. 98.3%, P = 1.000), while blood loss was lower in MILR group (200 vs. 350, P = 0.040) along with borderline faster functional recovery (3 vs. 4, P = 0.050). After a mean follow-up of 39 months, median DFS and OS and overall survivals resulted comparable between the two groups (33 vs. 36 months, P = NS; 51 vs. 63 months, P = NS)^[47].

In 2021, two meta-analyses of non-randomized retrospective comparative studies of MILR for ICC versus open were published. Both publications considered eight original researches published between 2016 and 2020 involving a total of 2,872 patients (552 MILRs and 2,320 open resections)^[48,49]. The baseline characteristics and perioperative outcomes were consistent between the two meta-analyses. MILR was characterized by lower transfusion rates (7.1% vs. 17.1%; P < 0.050), morbidity rates (20% vs. 32.6%; P < 0.050), and length of hospital stay (MD: -3.48 days, SMD: -0.40 days; P = 0.050) compared to open resection; no differences in operative time (MD: 1.6 min, SMD: 0.15 min; P > 0.050) and mortality (1.6% vs. 0.9%;P > 0.050) were found. MILR patients had smaller tumors (MD: −1.17 cm, SMD: -0.3 cm; P < 0.050) and underwent major resections less frequently (41.2% vs. 55.5%; P < 0.050); this is reasonably the explanation for higher radical resection rates for MILR (85.6% vs. 74.6%; P < 0.050). For oncological outcomes, recurrence-free survivals were comparable between MILR and open resection at both 3 years (HR: 0.97; P > 0.050; HR: 0.77; P > 0.050) and 5 years from resection (HR: 0.91; P > 0.050). Although no difference in 3-y overall survival was reported by Machairas et al. (HR: 0.9; P = 0.63), a significantly lower 5-year overall survival for MILR was shown by Regmi et al. (HR: 3.01; P < 0.001), albeit associated with fewer deaths due to tumor recurrence (22.3% vs. 35.4%; P < 0.050). Hence, the authors raise the question of the oncological adequacy of MILR for ICC and further evidence needed to address the oncological findings. In the same year, Ziogas et al. performed a survival meta-analysis based on reconstructed individual patient data regarding laparoscopic versus open hepatectomy for intrahepatic cholangiocarcinoma^[50]. The study was based on the same eight original researches and included one- and two-stage survival analyses. Using Cox regression, the authors showed no difference regarding OS (HR: 1.11; P = 0.70) and RFS (HR: 1.19; P = 0.47), thus concluding in favor of the oncological adequacy of MILR for ICC [Table 2].

Although surgical resection is the mainstay of treatment for both tumors and, in many cases, HCC and ICC can be found on postoperative histology in the same liver, differentiating between HCC and ICC is important for proper surgical planning. The main surgical peculiarity is the indication to perform locoregional lymphadenectomy. At present, lymph nodes retrieval in HCC is indicated only in case of documented metastases, for which the role of positron emission tomography/computed tomography scan with dual tracer ¹⁸F-FDG and ¹⁸F/¹¹C-choline holds promise for complete preoperative staging and operative decision^[57]. Instead, for ICC, a routine locoregional lymphadenectomy is supported by data on its favorable effect at least on staging accuracy and potentially also on curative goals and oncological prognosis^[39,58]. In particular, regarding the issue of satisfactory lymphadenectomy during MILR for ICC, it should be noted that although the mean number of retrieved nodes was not different, in all three meta-analyses, MILR resulted associated with lymph nodes dissection less frequently than open resection (47.8% vs. 61.9%; P < 0.05). Differences in the period of performance of surgery, in the attitude of the center and the surgeon, and in minimally invasive expertise may explain this difference. The development of minimally invasive skills is necessary for this tricky surgical step. Recent studies have shown that minimally invasive lymphadenectomy can be performed without affecting the accuracy of nodal dissection and can be possible even in case of unfavorable anatomical variations, disclosing some technical tips^[59-61].

The use of intraoperative fluorescence imaging in this setting is worth mentioning, as it may facilitate anatomical MILR and appropriate lymphadenectomy. Analyzing a small series of laparoscopic left hemi-hepatectomies, Zhang et al. recently reported that more lymph nodes were harvested in the indocyanine green group (7 vs. 3.5, P < 0.001), and the percentage of confirmed pathological lymph nodes was higher (75.0%, 66.7%-87.5% vs. 40%, 33.3%-50.0%, P < 0.001). They also found that indocyanine green staining was observed in all patients at stations 8 and 12, and in 75% and 92% of patients at stations 13 and 7, respectively, concluding that the indocyanine green fluorescence imaging system is feasible, safe, and effective to trace lymph nodes. It can be used to identify the patterns of regional lymphatic drainage. It can also help to define the extent of lymph node dissection in patients with ICC on a case-by-case basis^[62].

Similar to HCC, no studies have directly compared the outcomes of MILR for ICC performed in Eastern and Western countries. Data from systematic reviews and meta-analyses show no difference in indications and outcomes^[18,63]. This is an interesting and unexplored topic, considering that the risk factors and causes of chronic liver disease as well as genetics, are partly different in world areas and could underlie a possible difference in outcomes^[64,65]. The issue deserves a specific evaluation through dedicated studies and could be the subject of future multicenter investigations.

Current data support MILR for ICC when performed in high-volume hepatobiliary centers with a significant experience in minimally invasive liver surgery, as these procedures are still innovative. Few comparative studies are currently available; as such, the findings mentioned above need to be further investigated with larger, high‐quality comparative studies targeting both technical and oncological outcomes.

Figures 1 and 2 summarize the pros and cons of minimally invasive approaches for liver resection of HCC, ICC, and both.

Figure 1. PROS of minimally-invasive liver resections.

Figure 2. CONS of minimally-invasive liver resections.

PATIENT SELECTION AND DIFFICULTY SCorES For MILR

MILR is nowadays described as feasible for all liver cancers, type of resections, size and location. However, MILR for tumors located posteriorly and deep in the parenchyma has been acknowledged as complex since the early periods of application. In fact, the first worldwide consensus on MILR, held in Louisville in 2008, identified liver resections for lesions located in the peripheral segments (i.e., segments: 2-6) as the best indications for laparoscopy^[66]. To date, laparoscopic segmentectomies on posterosuperior segments (i.e., segments 4a, 1, 7 and 8) remain technically challenging despite advancements, as well as right posterior sectionectomies, caudate lobe resections and parenchyma-sparing resections. Compared to MILR on anterolateral segments, these hepatectomies result in greater blood loss, longer operative time, and higher morbidity and conversion rates^[67-69]. In any case, even if the localization of the tumor is an extremely determining factor for the feasibility, safety and reproducibility of MILR, the evaluation of the complexity should be multifactorial as there are various factors that significantly contribute to the overall difficulty, which should then be - ultimately - related to the experience of the surgeon. With this vision, some useful and easy-to-apply scoring systems for predicting MILR complexity have been developed and validated and are currently popular tools for clinical practice and safe implementation of MILR. Table 3 summarizes the existing systems with their characteristics and grading of complexity.

For HCC, all four scoring systems have been shown to perform well in predicting operative time, blood loss, major postoperative complications, and hospital stay^[72]. Of note, the Iwate scoring system has been specifically validated for application in patients with HCC, particularly for predicting conversion to open and long-term oncological results^[72,73]. On this topic, Lv et al. reported that the Iwate grading correlates with short- and long-term outcomes in HCC patients in view of significant differences in intraoperative (operative time P < 0.05; blood loss P < 0.05), postoperative (hospital stay P < 0.050) and 5-year disease-free and overall survival (both P < 0.050) among the four categories^[72]. Instead, for ICC, no specific validation of the difficulty scores has been studied to date.

CONCLUSION

HCC surgery is particularly suitable for the minimally invasive approach. Laparoscopic resections for HCC are associated with improved short-term outcomes (reduced rate of ascites and liver failure) compared with open surgery, without compromising long-term oncological outcomes. These features of MILR are a landmark change in the last two decades for the surgical treatment of this disease and are the basis for a possible expansion of surgical candidates in patients with advanced liver dysfunction (portal hypertension, severe cirrhosis), which is now also to be considered when developing treatment guidelines for HCC. Although still in the phase of exploration, the use of minimally invasive surgery for ICC seems feasible and safe and short-term patient outcomes appear to benefit, but more evidence is needed to validate the long-term oncological outcomes compared to open. Should it be demonstrated that the short-term benefits are accompanied by non-inferior long-term oncological outcomes and the reproducibility of an adequate locoregional lymphadenectomy, MILR could be endorsed by expert guidelines as a standard treatment in centers with acquired expertise in the technique. Prospective trials leading to improvement of life expectancy after surgery both in HCC and ICC would be the optimal tool for research methodology; despite the difficulty, they should be encouraged.