Introduction

Over the last 20 years, laparoscopic surgery has evolved to become the approach of choice for many abdominal procedures [1]. Excellent results in terms of hospital stay and quality of life has permitted this technique to become the gold standard for cholecystectomy and other procedures. Hepatic surgery has evolved dramatically with an improved understanding of the anatomic segments of the liver; enhanced imaging by CT and MRI scans; improved anesthesiology, critical care, postoperative nursing, and physical therapy; and technological advances and modifications in laparoscopy [2] for benign and malignant tumors. With the exception of hepatic adenoma (HA), surgical indications for hepatic hemangioma (HH), and focal nodular hyperplasia (FNH) of the liver remain controversial [3].

Laparoscopic liver surgery for benign hepatic tumors was first reported in 1991 [4]. More recently, increased experience in laparoscopic liver surgery and the contribution of improved technology have fuelled enthusiasm for this surgical approach. Increasing numbers of reports have now established that despite occasional longer operating times, laparoscopic liver surgery is associated with reduced blood loss, reduced postoperative morbidity, and shorter hospital stays [5].

As reported for other conditions, the advent of laparoscopy has increased the number of unexplained procedures with a negative impact on health care cost [6].

The aim of this study was to analyze the risk of an increased surgical indication rate in patients with benign tumors of the liver since the development of laparoscopy.

Patients and methods

An extensive search of relevant literature restricted to English, Italian, and French languages was performed using MEDLINE (PubMed), Google Scholar, and The Cochrane Library. The date of the last electronic search was March 31, 2011, and the period included 1970 to 2010.

The keywords used for the search were: ‘benign liver tumor’, ‘laparoscopy’, ‘hepatic resection’, ‘liver resection’, ‘hemangioma’, ‘FNH’, ‘focal nodular hyperplasia’, and ‘adenoma’. These keywords were used individually or with the help of the Boolean operator ‘AND’.

Inclusion criteria were patients with a benign tumor of the liver (only hepatic hemangioma, focal nodular hyperplasia, and hepatic adenoma) who underwent elective surgery without an indication for emergency surgery (e.g., hemorrhage or rupture). Articles that did not clearly meet the inclusion criteria were excluded at this stage.

All titles and abstracts were screened, and those related to the theme of this review article were selected. We collected all considered articles, number of patients, number of benign liver tumors, number of patients who underwent operations, and type of surgical procedure (open or laparoscopic surgery). We considered patients >18 years of age to ensure analysis of only adult patients. Age and sex were not considered for the present analysis because they did not influence the search or, consequently, the results.

All articles were then divided into two groups: group I included all manuscripts reporting patients who underwent an open procedure between 1971 and 1990, and group II included all manuscripts reporting patients who underwent an open or laparoscopic procedure between 1991 and 2010.

The amount and percentage of HH, FNH, and HA diagnosed, submitted to surgery, and not treated surgically were analyzed in the two groups. We divided all group II articles into two subgroups. In subgroup IIA, patients treated by an open or laparoscopic procedure between 1991 and 2000 were analyzed. The remaining patients treated by an open or laparoscopic procedure from 2001 to 2010 represented subgroup IIB. The percentage of procedures performed laparoscopically in the two groups was compared using a chi-square test with Yates’ correction.

Results

A total of 10,584 reports were found and reviewed (Fig. 1). After examination of all titles and abstracts, 5,712 articles were immediately excluded as not pertinent on the basis of the title and aim of the manuscript; 3,975 articles were excluded because they were identified twice in two databases; 493 articles were excluded because using the Boolean operator, they involved primary liver tumors without any description of a benign liver tumor; and 128 articles were excluded for the same reason but concerned secondary liver tumors.

Fig. 1
figure 1

Algorithm used to screen the literature

Full size image

Of the remaining 276 articles, 49 were excluded due to the absence of a description of surgery, and 23 articles were excluded because they were case reports. The remaining 204 articles were fully reviewed, and 109 were excluded because they did not adhere to our protocol. At the end of the search, only 95 articles addressed all inclusion criteria and were used for the present study (Fig. 1).

A total of 5,480 patients with benign lesions of the liver were found from the final research of the literature articles meeting all criteria for the present study. A total of 1,071 (19.5 %) patients did not undergo surgery and were therefore excluded from the work, and 4,409 (80.4 %) patients underwent a surgical procedure and were analyzed in the present work. These included 2,492 (45.6 %) patients with HH, 942 (17.1 %) with FNH, 610 (11.1 %) with HA, and 1,436 (26.3 %) with other types of lesions. The majority of liver resections were performed in 1,661 (37.6 %) patients with HH, followed by 825 (18.7 %) with FNH, and 540 (12.2 %) with HA. The remaining 1,383 (31.5 %) patients had other benign liver lesions that were not considered for the present work. Thus, a total of 3,026 hepatic resections were ultimately considered for the present work (Fig. 2).

Fig. 2
figure 2

Algorithm of patients undergoing hepatic resection. Patients are divided according to the type of period and method (open or laparoscopy surgery) used for their treatment

Full size image

In the first group (group I [GI]) of articles, published between 1971 and 1990, we analyzed 13 articles, among which 545 patients were considered for the present study. Of these patients, 393 (72.1 %) underwent traditional surgery and 152 (27.9 %) were not submitted to surgery. There were 213 (54.2 %) patients with HH, 34 (8.7 %) with FNH, 71 (18 %) with HA, and 75 (19.1 %) with various kinds of lesions excluded from the present study (Fig. 2). Thus, in GI, we definitively analyzed only 318 patients affected by 213 (67 %) HH, 34 (10.7 %) FNH, and 71 (22.3 %) HA and undergoing an open surgical procedure (Table 1) [7–19].

Table 1 Articles of the GI group published between 1971–1990. Only open technique is used for the treatment of benign tumors of the liver in this group of patients
Full size table

In the second group (group II [GII]) of articles, published between 1991 and 2010, we analyzed 82 articles, among which 4,935 patients were affected by benign tumors. A total of 4,016 (81.4 %) of these patients underwent surgical procedures, and the remaining 919 (18.6 %) were not treated surgically. There were 1,448 (36.1 %) patients with HH, 791 (19.7 %) with FNH, 469 (11.7 %) with HA, and 1,308 (32.5 %) with different hepatic lesions not useful for the present study (Fig. 2). Therefore, in this group, 2,708 patients affected by 1,448 (53.5 %) HH, 791 (29.2 %) FNH, and 469 (17.3 %) HA and undergoing a surgical procedure were studied (Table 2) [20–101]. In particular, 2,112 (78 %) patients were treated with open surgery and 596 (22 %) patients were treated with laparoscopic surgery.

Table 2 Articles of the GII group published between 1991–2010. Open and laparoscopic technique were used for the treatment of benign tumors of the liver in this group of patients
Full size table

Analysis of the two groups showed that the number of benign lesions of the liver diagnosed in GI was less than that in GII with a ratio of 1/9 (545 vs. 4,935, respectively). The increase in the diagnosis of benign lesions of the liver resulted in an increase in surgical procedures in GII compared with GI (4,935/4,016 vs. 545/393, 81.3 % vs. 72.1 %, respectively). The total number of diagnosed benign tumors of the liver vs the total number of only HH, FNH, and HA in GI was higher than that in GII (545/439 vs. 4,935/3,605, 80.6 % vs 73.1 %, respectively).

The total number of HH, FNH, and HA treated with surgical procedures related to the total number of HH, FNH, and HA diagnosed in GI was lower than that in GII (318/439 vs. 2,708/3,605, 72.4 % vs. 75.1 %, respectively).

Analysis of the number of surgical procedure for each kind of tumor (HH, FNH, and HA) in relation to the total number of surgical procedure for all these tumor in GII related GI showed a decrease for HH (1,448/2,708 vs. 213/318, 53.3 % vs. 67 % respectively) and HA (469/2,708 vs. 71/318, 17.2 % vs. 22.3 %, respectively) and an increase for FNH (791/2,708 vs. 34/318, 29 % vs. 9.7 %, respectively).

The last analysis concerns the two subgroups. In subgroup IIA (GIIA), a total of 31 articles were published between 1991 and 2000, and 917 patients reportedly underwent a surgical procedure (Table 3). In this group, 872 (95.1 %) patients underwent an open surgery and 45 (4.9 %) underwent a laparoscopic procedure. In particular, 434 (49.8 %) patients with HH were treated with open surgery and 16 (35.6 %) with HH were treated with laparoscopic surgery, 286 (32.8 %) patients with FNH were treated with open surgery and 16 (35.6 %) with FNH were treated with laparoscopic surgery, and 152 (17.4 %) patients with HA were treated with open surgery and 13 (28.8 %) with HA were treated with laparoscopic surgery (Fig. 2).

Table 3 Articles of the GIIA sub group published between 1991–2000. Open and laparoscopic technique were used for the treatment of benign tumors of the liver in this group of patients
Full size table

In subgroup IIB (GIIB), a total of 51 articles were published between 2001 and 2010, and 1,791 patients reportedly underwent a surgical procedure (Table 4). In this group, 1,240 (69.2 %) patients underwent an open surgery and 551 (30.8 %) underwent a laparoscopic procedure. In particular, 814 (65.6 %) patients with HH were treated with open surgery and 184 (33.4 %) with HH were treated with laparoscopic surgery, 259 (20.9 %) patients with FNH were treated with open surgery and 230 (41.7 %) with FNH were treated with laparoscopic surgery, and 167 (13.5 %) patients with HA were treated with open surgery and 137 (24.9 %) with HA were treated with laparoscopic surgery (Fig. 2).

Table 4 Articles of the GIIB sub group published between 2001–2010. Open and laparoscopic technique were used for the treatment of benign tumors of the liver in this group of patients
Full size table

The percentage of patients treated with a laparoscopic procedure was less in GIIA than in GIIB (4.8 % vs. 30.8 %, respectively). A chi-square test with Yates’ correction gave a P value of <0.001.

Discussion

Laparoscopy is slowly but definitively changing the course of surgery. It has benefits of a fast recovery and rapid return to activities because postoperative pain is less or nonexistent compared with open surgery. As for the majority of surgical diseases, hepatic surgery has been gently changed by laparoscopy during the last 20 years, and a great number of hepatic resections are now performed by laparoscopy.

The manuscripts in the present search were unequal during the 40 years included in the study in terms of the two groups of populations studied. This is due to the natural evolution of hepatic surgery: in the first two decades, the initial experience with few surgical teams has been progressively replaced with many teams, affirming this kind of surgery worldwide. The change from open to laparoscopic hepatic surgery by many of these teams or new laparoscopic teams has influenced hepatic surgery in the last two decades. This surgical evolution perfectly corresponds with the literature. The first few scientific reports concerning the initial experience of a few groups were replaced with an increasing number of scientific manuscripts corresponding to an increase in open hepatic surgery teams until the appearance of laparoscopic teams, which resulted in a gradual increase in not only laparoscopic surgeries, but also the related scientific articles.

This is the reason why the last period of 20 years have been divided in two subgroups; in the first period only few groups of very skilled hepatic surgeons and at the same time excellent laparoscopists have performed this surgery. On the opposite the second period (the last 10 years) represents the enormous spread of this techniques performed not only by hepatic surgeons devoted to laparoscopy but also by laparoscopic surgeons (frequently general surgeons) initiates to hepatic surgery.

Liver resection represents the best treatment for a variety of malignant and benign hepatic tumors, and estimation of risk factors affecting the early outcome after hepatic resection is a goal shared by all high-volume centrals specialized in hepatobiliary surgery [102]. To achieve this goal, the right indication for each kind of hepatic tumor is mandatory to avoid unnecessary surgical procedures with related morbidity and mortality.

After the introduction of laparoscopic hepatic resection, many authors have detected a dramatic increase in the amount of surgical procedures performed in general hospitals [103]. Nguyen et al. used a laparoscopic approach in 25 % of cases in the last 6 years [2]. Koffron et al. shifted their practice from 10 % minimally invasive liver resections in 2002 to 80 % liver resections in 2007 as long as the patient met certain safety and oncologic requirements [103].

Despite the limitations and disadvantages of laparoscopic liver resection, which include a significant learning curve, bleeding that is more difficult to control laparoscopically, inadequate assessment of the liver for additional lesions, and increased risk for gas embolism, the increase in laparoscopic hepatic resections has been maintained; therefore, the explanation for this trend could broaden the indications for performance of such a technique. Surgical laparoscopic procedures are easy if they involve the anterior segments of the liver and are performed for benign lesions of the liver. In our review, among the manuscripts that reported these data, the majority of resections were minor. We can postulate that because all of the initial published experiences of laparoscopic hepatic resections were based on limited or minor hepatic resections, these procedures are easier to perform and they are chosen by novice laparoscopic teams because of decreased initial difficulties [104].

In the 1970s, ultrasonography used for diagnostic purpose permitted an increase and refinement of benign hepatic tumor diagnosis [104]. Together with a better knowledge of the natural course of these tumors, this helped decrease the total number of procedures in the second period permitting at the same time the avoidance of unnecessary liver surgery in asymptomatic patients. In fact, liver surgery for benign liver tumors may relieve complaints in a high percentage of symptomatic patients (80 %). However, in many patients, symptoms persist after resection of the tumor, and surgery-related complications might occur [51] so the right indications for surgery is mandatory.

MRI may be helpful when the diagnosis is dubious, while percutaneous biopsy is generally avoided. With this approach, most patients can be safely observed [54], but despite what, according to the increase in diagnosis of benign lesions of the liver, there was an increase in surgical procedures for such lesions. As in our study, Buell et al. found that the most common resected benign hepatic tumors were HH, followed by FNH and HA [84]. This was probably due to a better knowledge of the clinical course of patients with benign tumor of the liver, which led to surgeons operating on these patients to complete their learning curve.

The surgical indications for HH, FNH, and HA are reported in Table 5. In the majority of patients, HH remain asymptomatic and are incidentally discovered during a surgical procedure or imaging studies for unrelated problems. In GII of the present study, the number of resected HH was considerably decreased because of an improved understanding of the diagnosis, prognosis, and correct indications for surgery.

Table 5 Indication for surgical treatment of patients with benign tumors of the liver
Full size table

Ibrahim et al. found that in patients with FNH, the main indication was suspicion of malignancy. This not surprising because it is sometimes difficult to differentiate hepatocellular carcinoma from FNH by investigation [83]. However, the proportion of symptomatic FNH patients has remained stable or even decreased in GII because more and more cases of FNH are incidentally detected; thus, a steadily increasing number of laparoscopic liver resections are performed in asymptomatic patients.

The continuous improvement in diagnostic techniques, particularly MRI, has restricted the surgical indications for HA to the non-classic or mixed hyperplastic adenomatous form of FNH [105], for which the procedure has not only a curative goal, but also the need for a precise pathologic definition [106]. Our review shows that in GII, there was a reduction in the proportion of patients with HA who underwent a surgical procedure compared with that in GI.

But despite these results the global analysis of the GIIB shows a significative increase in laparoscopic hepatic resections for all three kind of tumors and this seems unjustified in a period in which the right indications for benign hepatic tumors of the liver were quite well established. This is also confirmed by the consensus conference in Louisville KY, USA called for an international position on laparoscopic liver surgery said stated that (1) incidental findings of benign asymptomatic liver lesions has become common, (2) HH and FNH can be diagnosed in most cases by imaging alone and rarely require surgery, (3) HAs are recognized to possess a potential for bleeding and malignant degeneration, and, most importantly, (4) the consequences of an adverse event are magnified when a procedure is performed for asymptomatic benign lesions [1].

Formal general surgical residency training in laparoscopy techniques began in the early 1990s and, for several years, was largely limited to cholecystectomy. Subsequently, in the 1990s, this training was extended to include other general surgical procedures such as hepatic resection [103]. Hepatic resection was one of the last frontiers of the laparoscopic approach because of the particular expertise required to perform this technique. In effect, at the beginning of this translation from open to laparoscopic, only hepatic surgeons performed these procedures.

Training of the surgeons is essential because we have seen a 300 % increase in demand for operations. However, this must be evaluated based on the potential increase in operative costs, which can balance out a reduction in postoperative hospital costs [107]. In the present study, some authors reported an increase in laparoscopic lobectomy related to open surgery (28 vs. 8, respectively; P = 0.001) [107].

In these authors’ opinion, the indication did not change even when laparoscopy was requested. However, knowledge of possible adverse laparoscopic effects on malignant tumors has probably pushed toward operation on more codified benign tumors [107].

The initial factor in learning and moving toward laparoscopic hepatic lobectomy is to choose patients requiring wedge resections of superficially or peripherally located neoplasms and left lateral hepatectomies.

Small, focal, localized tumors on anterolateral segments (segments II–VI according to the Couinaud classification) are typically considered for easier resection [108]. The majority of published manuscripts concerning the initial experiences of a novice laparoscopic team concern limited or minor resection [45, 63, 67]. The increase in laparoscopic procedures in GII was probably caused by two factors. For benign tumors of the liver, no radical margins are requested [109], and wedge or limited resection of the anterior segment of the liver can be easily performed with the double advantage of easy resection and facilitation of training [110].

Moreover, normal parenchyma is more manageable than is cirrhotic liver and is another reason why benign tumors of the liver can be treated more easily [109].

There is no reason for modification of the management of patients suffering from benign liver tumors after the introduction of a laparoscopic procedure. Surgical indications for removal of these tumors should be based on their natural history and the ability of imaging techniques to ensure a precise diagnosis of the type of tumor [60]. Adequate selection of patients and liver tumors is a key factor for successful laparoscopic resectional surgery [60]. The procedure should be performed by surgical teams experienced in hepatobiliary and laparoscopic surgery [60]. Studies have shown that there are no financial disadvantages to the laparoscopic approach because the added costs of disposable equipment or devices in the operating room were offset by shorter operative times and lengths of stay [2]. Further reduction of abdominal wall damage and cosmetic advantages of the laparoscopic approach represent a clear benefit in patients with benign tumors of the liver. Of course, if indication of laparoscopic surgery is not mandatory, the procedure itself represents an important increase in health care costs.

No evidence is currently available to support or refuse the indication for laparoscopic surgery in elective patients with benign liver lesions. However, the present analysis shows that in the last 20 years, there has been an increase of 26 % in laparoscopic procedures for benign tumors of the liver.

A limitation of the present study concerns the factors that affect the apparent frequency of the procedure. This can be caused by referral bias of the patients affected by hepatic benign tumors. Luning showed that ultrasound had an accuracy of 69 %, CT had an accuracy of 73 %, and MRI had an accuracy of 80 % in demonstrating the type of lesion [110]. Another bias concern the absence of articles against laparoscopy, infact all published manuscripts discuss the advantages of this procedure. Thus, it is very difficult to analyze the real causes of the increase in laparoscopic procedures. Last bias may concern the change in incidence of tumors studied: the relationship between the tumors and one of their causes (oral contraceptives) can be considered because this treatment has increased during the last several years and may have also influenced the incidence of this tumor [110]. Finally, the postoperative course of the benign tumors with a normal liver is now well codified and does not need any particular expertise. Thus, in terms of other benign diseases, the patient can be discharged very quickly, resulting in personal satisfaction and appreciation for the technique.

In conclusion, the improvements in diagnostic techniques during the last few decades have permitted diagnosis of more benign tumors of the liver.

The indications for surgery were codified, and the surgical procedures remained stable for a period of time before the advent of laparoscopy; with the advent of this technique the hepatic resections for benign tumors of the liver has increased inexplicably.