The first MILRs reported at the beginning of the 1990s were basically wedge resections of peripheral lesions (18). Subsequently, anatomic resections such as left lateral sectionectomy were performed (19). The first major laparoscopic hepatectomy series was reported in 1997 by Hüscher et al. (20) using hybrid procedures for right-sided resections.
Although it has several theoretical advantages, only a small percentage of liver resections are actually performed by minimally invasive surgery. A recent French national database study showed that 15% of liver resections were performed through minimally invasive approach (21). Similarly, Kim et al. (22) showed that less than 10% of all liver resections for benign lesions in the United States were minimally invasive.
Currently, minor laparoscopic resections in anterolateral segments and left lateral sectionectomy are considered the gold standard approach in many specialized centers (3, 8, 23). However, resections of bilateral lesions, nodules in posterosuperior segments or in central locations of the liver (segments 1, 4a, 7, and 8), and major hepatectomies are still challenging (24, 25). In fact, technical demands have limited major hepatectomies to highly-skilled surgeons in referral centers (10, 26). Concerns during anatomical right liver resections are related to liver mobilization from the inferior vena cava, inflow and outflow control, and a large parenchymal transection area. Moreover, the learning curve for MILR can reach 45-75 procedures (11, 27).
Technical limitations of laparoscopic major resections were depicted in a recent survey including 27 specialized centers. While minimally invasive approach was used in 61.8% of left lateral sectionectomies, this percentage decreased to 24.8% for major hepatectomies (28).
Robot-assisted surgery has been increasingly employed as an alternative to laparoscopy for MILR, mainly in complex and major liver resections (29, 30). Despite potential advantages, most of the available evidence demonstrated similar results between laparoscopic and robotic liver resections (5, 31, 32).
Few studies were addressed to study the results of minimally invasive major hepatectomies (10, 26, 33). Only recently observational studies with high methodological quality have been published comparing open and minimally invasive major resections (34, 35). Takahara et al. (36) using PSM showed advantages in terms of blood loss, length of hospital stay, and complications with the laparoscopic approach.
We observed a conversion rate of 5.4%, lower than observed in other series raging from 9% to 42%. Cipriani et al. (37) showed an 11% conversion rate in a European multicenter study. Similarly, Kasai et al. (33) observed a 17.7% conversion rate in a recent meta-analysis of individual data.
MILR is frequently associated with a longer operative time (34, 35). However, we did not find any significant difference in operative time for patients undergoing MIRH. This finding can be explained by the increased experience with MILR, showing that the learning curve was overcome and surgical steps have been standardized to entail a significant reduction in operative time (38, 39). In fact, were included in our study patients that underwent hepatectomy between 2013 and 2018, after overcome the learning curve with MILRs. Minimally invasive liver surgery program started at the University of São Paulo in 2005 and at the Diaconesses Croix Saint Simon Hospital in 2010. The 2 centers altogether have performed more than 550 minimally liver resections.
In accordance with previous studies (34, 36), we observed a significantly lower blood loss in the MIRH group. Factors that may have influenced this reduction are the development of new energy devices for liver transection, the image magnification afforded by laparoscopy, the pneumoperitoneum, and the widespread use of linear staplers for controlling hepatic pedicles and large vessels (1).
Several authors found a reduction in perioperative complications in the MILR group (36, 37). However, in our study the overall morbidity (our primary end-point) presented a non-significant decrease (35.1% vs. 53.3%, P=0.09), probably related to the small sample size of our casuistry. Although non statistically significant, this finding appears to be clinically relevant. Our data showed a decrease in Dindo-Clavien I-II complications (13.5% vs. 35%, P=0.03), in accordance with the findings of a recent meta-analysis (33).
The reduction of hospital stay is an outcome frequently attributed to minimally invasive surgery (2, 36). We observed a 2-day reduction in the MIRH group. Although not statistically significant, we considered this a consistent clinical benefit for patients subjected to major liver resections.
Concerning the oncological outcomes, there was no increase in R1 resections in MIRH group. This data is in accordance with other studies that found similar R0 resections, and wider margins associated with MILRs when compared to OLRs (2, 40).
Few comparative studies assessed the long-term results of minimally invasive resections (33, 41). The fear of inferior oncological results in patients undergoing MILRs was not demonstrated by the available studies (33, 35). Similarly, in our study the OS rate in MIRH group was not inferior when compared to isolated ORH group. The same finding was observed in subgroups of patients operated for primary liver tumor and liver metastases.
Our study has the classical drawbacks of any retrospective analyses. For this reason, we focused only in right anatomical liver resections. Moreover, in order to minimize selection bias we excluded complex or associated procedures such as two-stage hepatectomies, hilar cholangiocarcinoma and synchronous resections. Finally, we used the PSM and observed that after matching, both groups were homogeneous in the main clinical and surgical characteristics. PSM was used to limit observed confounding factors, however it is important to state that PSM does not allow us to control for possible selection bias. Also, by introducing secondary outcomes and performing other comparisons, we could have increased the alfa risk and decay our results. Altogether, these points may limit the extent to which we can generalize our findings.