An independent association of the number of resected segments with postoperative complications and mortality rate is clearly described. The main series focusing on trends in perioperative outcomes after liver resection over the decades have shown a significant reduction of mortality rate. Nevertheless, despite better patient selection and perioperative management, the rate of postoperative mortality after major or extended resection did not change over the years. Therefore, we can suppose that much of the decrease in mortality was probably related to the intensely diffusion of PSLR. A recent study on 3875 patients with CLM treated from 1993 to 2012 at Memorial Sloan Kettering [2] reported a significant decrease of the median number of resected segments over the years (from 4 to 2), with a simultaneous decreased mortality rate from 5.2 to 1.6%. Interestingly, these results were not confirmed in the subgroup of patients treated with major liver resections. Many authors reported that PSLR was also associated to a significant reduction of postoperative morbidity. In particular, Gold et al. [3] showed at multivariate analysis that the rates of liver-related complications were strictly correlated to the number of resected segments.

In fact, postoperative liver function is well preserved in PSLR with postoperative serum bilirubin level significantly lower and prothrombin time significantly higher with respect to patients who received major resection [4]. These results concerning liver function are even more interesting if we consider the adverse effects of neoadjuvant chemotherapy on liver parenchyma. It’s important to note that the main cause of death after liver resection is still liver failure. So it is crucial to prevent this complication for which no efficient therapies are available. Finally, Kingham et al. [2] reported lower estimated blood loss and transfusion rate, as fewer segments were resected. These data are important given the negative impact of both these factors on long-term results.

The development of PSLR is based on the assumption that it meets the principles of oncological surgery. Per definition, PSLR is more likely to be associated with smaller margins. It is known that intrahepatic micrometastases via portal branch are uncommon in case of colorectal liver metastases, as reported by the histopathological study performed by Yamamoto et al. [5]. In addition, we previously reported [6] that the width of a negative surgical margin does not affect the risk of local recurrence or survival. A recent propensity score-matched analysis of nearly 3000 CLM hepatectomies validates these data [7]. In confirmation of oncological suitability of PSLR, many studies reported similar disease-free survival and liver recurrence rate in patients treated with or without conservative liver resection. One advantage of leaving the maximal amount of liver parenchyma is to enhance the possibility of repeat hepatectomy in case of liver recurrence. From 60 to 70% of patients undergoing liver resection for CLM will develop recurrence of the disease [8]. Of these, one third will have isolated liver recurrence. It has been demonstrated that patients who undergo a second or a third liver resection for recurrent liver metastases may have long survival. According to this notion, some studies reported better overall survival for patients treated with PSLR compared with major resections, due to a higher likelihood of undergoing salvage hepatectomy for recurrence. One can argue that the majority of these reports compared heterogeneous groups of patients without matching tumor number or size, which makes the interpretation of the data difficult. Nevertheless, Mise et al. [9] recently compared long-term results of patients with a single CLM less than 3 cm treated with (156 patients) or without (144 patients) PSLR. In this cohort of patients with uniform tumor characteristics, the authors confirmed better overall survival (72.4% vs 47.2%, p = 0.047) and higher redo-resection rate (68% vs 24%, p < 0.01) in PSLR group.

IOUS has a key role in modern hepatic surgery not only to better stage the disease, but above all as guidance to resection [10]. The extensive use of IOUS allows to maximize the parenchymal sparing of healthy liver tissue. In fact, conservative liver resections in case of metastases deeply located can result in thin surgical margins, with the risk of exposing the tumor during the parenchymal transection. IOUS reduces this risk, as it makes it possible to precisely identify the site and the number of lesions, showing the relationships between the metastasis and the main vascular structures that can be followed in real time. Furthermore, IOUS permits to check the correct plane during parenchymal transection.

PSLR is a philosophy rather than a surgical technique that encompasses a wide range (countless actually) of liver resections, ranging from small non-anatomical wedge resection, to complex atypical non-anatomical resections (Fig. 16.2), to segmentectomies or subsegmentectomies. The crucial feature according to which type of resection is more appropriate is the need to sacrifice liver vessels, both glissonian pedicles and hepatic veins. Therefore, it is clear that a meticulous IOUS exploration of the liver is the necessary prerequisite for modern liver surgery. It is very important to be methodical in the IOUS exploration, in order to perform a complete and accurate examination. We recommend performing two standardized explorations. The first scan is for liver anatomy assessment, and is performed in a systematic manner, i.e., starting from the hepato-caval confluence and hepatic veins, than moving to portal bifurcation, left and right segmental pedicles till the third order of division. The second systematic exploration is for IOUS staging. The surgeon explores the parenchyma looking for new lesions, and precisely locates the metastases in the three-dimensional map of liver vascular structure he just explored. During this step, the tumor vascular relationships are carefully assessed. IOUS makes it possible to measure the distance between a metastatic nodule and a vessel or, in the case of adhesion to the vessel, to define the longitudinal and circumferential extent of the contact. Vessels in contact with metastases have to be deeply explored for signs of infiltration such as tumor thrombi, or direct infiltration with endoluminal growth. Other indirect signs of infiltration such as uneven vessel wall profile, disappearance of the hyperechoic venous wall (Fig. 16.3), or biliary dilatation around glissonian pedicles (Fig. 16.4) can be identified. According to all these elements, the surgeon will decide which vessel to spare and which to cut, thus planning the extent of the resection. If vascular infiltration is suspected, the vessel should be ligated and sectioned in order to allow radical resection. Otherwise, the lesion can be dissected from the vessel, even if a thin surgical margin is obtained. However, in the case of infiltration of hepatic veins, when accessory veins or communicating veins are present, it may not be necessary to remove all of the liver parenchyma drained by that vein. In this setting, knowledge of the liver inflow and outflow by IOUS is fundamental for determining the feasibility of surgery. In fact, a third exploration is needed. It is the real IOUS guidance to whatever resection. As opposed to the first two, this exploration is focused on the lesion to be removed, and is performed after appropriate liver mobilization that can be minimal or extensive, with the liver lifted by the left hand of the surgeon. Relationships between tumor and surrounding vessels are explored again in order to visualize and draw on the liver surface with cautery hepatic veins, portal pedicles and their branches, directly involved or just surrounding the lesion (Fig. 16.5). Vessel can be scanned in the longitudinal plane and the brief impression on the glissonian capsule left by of the probe mildly pressed can be marked with cautery. When the vessel is cross-scanned, it can be targeted with the artifact of the tip of the electrocautery slipped beneath the US probe and then properly marked. A series of cross-section marks can be joined to draw the longitudinal axis of the vessel. After the map of the area has been traced, the lines of transection are drawn accordingly. During parenchymal transection, IOUS allows monitoring of the correct surgical plane in order to maintain an adequate surgical margin and avoid lesions to major vascular structures. Finally, IOUS can be used to assess the correct drainage of remnant liver. Inflow and outflow are evaluated to detect ischemic or congestive areas. At the end of the operation, the raw cut surface of the liver can be visualized by IOUS, allowing the detection of any remaining lesion. The specimen itself can be explored outside the operative field to check the lesions and the surgical margin.

Resections of small, superficial metastases often do not require pedicle dissection or ligation. Nevertheless, the importance of IOUS should not be underestimated. In the first place, to ascertain the distance of vessels and that indeed none has to be ligated. Then, IOUS helps to choose the proper distance between the section line and the metastasis projection (Fig. 16.6). It is in fact important to allow an adequate distance, not only to get a negative margin in the lateral aspect but especially in the deep plane. The transection route has to reach the deepest point beneath the lesion with an even angle and adequate margin. Beginning the transection too close to the lesion would require an almost vertical route that would make the resection difficult and unsafe. In fact, it is not uncommon to achieve a positive margin for “easy” resection. Besides, limited resections does not always mean easy resections. Multiple bilobar atypical resections, possibly associated with segmentectomies or subsegmentectomies, are often lengthy and demanding procedures inconceivable without IOUS planning and constant guidance (Fig. 16.7). When a glissonian pedicle is involved and cannot be spared, it can be checked by IOUS, precisely targeted and ligated in the dissection, even though in many cases the extent of the resection is decided upon the involvement of hepatic veins, whose ligation could impair the outflow of (theoretically) a whole liver sector. When IOUS shows no sign of infiltration or limited contact, lesions can be detached during dissection, thus sparing the vessel involved. IOUS guidance makes it possible to reach the proper vein with a correct angle and to guide the dissection exposing the spared hepatic vein (Fig. 16.8). In case of focal infiltration, hepatic veins can be partially resected en bloc with the tumor, and reconstructed with direct suture or with a patch (Fig. 16.9). Cases with multiple lesions, with various glissonian and hepatic vein involvements, can often still be treated with limited (meaning non-formal major hepatectomies) resections, unachievable without IOUS guidance (Fig. 16.10). Concerning CLM, an anatomic segmentectomy is required when a glissonian pedicle has to be sacrificed. In that case, the whole segment fed by the pedicle has to be removed in order not to leave ischemic parenchyma. The sacrifice of a hepatic vein can also require the resection of the drained segments. The difficulty of segmentectomies relies on the lack of landmarks on the liver surface to guide resection. Over the years, many methods have been proposed for liver segment identification. The first procedure was described by Makuuchi et al. [11] for anatomic resection for hepatocellular carcinoma, and consists of the puncture of the portal branch involved and injection of indigo carmine. As a result, the stained area becomes visible on the liver surface and can be marked by electrocautery. Although the dye staining is the most accurate method of segmental or subsegmental identification, it is not easily reproducible, and if a wrong pedicle is punctured, a mistaken area is stained, making it very difficult to identify the right segment once more. To apply the precision of staining in a reversible fashion, Torzilli et al. [12] proposed the compression of the portal branch technique. It consists in the IOUS identification of the feeding portal branch of the concerned segment. The branch is compressed between the IOUS probe and the finger, inducing a transient ischemia of the distal parenchyma. The area can then be marked by electrocautery. Another technique proposed by Machado et al. [13], first for left segments and then extended to almost all segments, is the intrahepatic glissonian approach. This technique requires small liver incisions according to anatomic landmarks such as the Arantius’ and round ligaments to isolate second- or third-order pedicles. After the ligation, transection can follow the ischemic line on the liver surface. Several segmentectomies can be safely performed with extrahepatic pedicle control, such as left liver segmentectomies (Sg2 and Sg3) or subsegmentectomies (Sg4a and Sg4b), as well as right sectorectomies: Sg6–7 and Sg5–8 (Fig. 16.11). The inferior pedicles of Sg4 can be dissected on the right side on the round ligament and ligated. The discoloration on the glissonian surface then drives the resection. In the same way, the pedicle of Sg2 and 3 can be managed on the left side of the ligament. The dissection of the right side of the hilum allows the identification of the portal branch and artery of the posterolateral and anteromedial sector. Their correct identification can be ascertained by temporary clamping and with IOUS with color Doppler showing no inflow in the proper segments. After the ligation, the ischemic line guides the transection. Nonetheless, continuous IOUS control of the transection is helpful. Concerning mono segmentectomies or subsegmentectomies, a pure IOUS-guided resection is feasible with a technique not different (and sometimes easier) from a large atypical resection. The lateral landmarks of a liver segment (course of the proper hepatic veins) are visualized and marked, as well as the glissonian pedicle. The parenchymal transection is carried out and the glissonian pedicle is reached under IOUS control. Its proper identification before ligation can be confirmed by the hooking technique [14]. The ligation of the pedicle causes an ischemic demarcation line that together with the anatomical IOUS landmarks allows the achievement of the resection. Tumor infiltration of a hepatic vein close to the caval confluence has for years entailed a major hepatectomy, namely a right hepatectomy in case of right hepatic vein infiltration and central hepatectomy for middle hepatic vein involvement. This is no longer acceptable in many cases; in fact there are several settings that allows the blood outflow from the territory of a closed hepatic vein. The first instance is the presence of an accessory hepatic vein clearly demonstrated at the preoperative workup and confirmed at IOUS. The most typical case, as first described by Makuuchi et al. in 1987 [15], is the bisegmentectomy of segment 7–8 with ligation of the right hepatic vein, thanks to the presence of a right inferior hepatic vein providing blood drainage of segment 6. Similarly, when the ligation of hepatic vein is required, it is always necessary to rule out the presence of vicarious accessory branches of the adjacent hepatic veins, such as branches of the middle hepatic vein draining Sg6 or branches of the left hepatic vein draining Sg4. In other cases, neither accessory vein nor communicating veins between adjacent sectors can be demonstrated preoperatively. Nevertheless, communicating veins can be identified intraoperatively thanks to color Doppler, especially through the latest high-resolution ultrasound colour flow modes that allow the identification of low flow vessels (Fig. 16.12). During the operation, as part of the anatomical exploration, color Doppler is used to check for the presence of communicating veins between the hepatic vein to be resected and the adjacent hepatic vein. If no communicating veins can be identified, a second exploration is performed after liver mobilization, dissection, and clamping of the hepatic vein that will eventually be resected. This maneuver can allow the perfusion of small communicating veins that are almost void in native state. Another method for ascertaining vicarious drainage was described by Sano et al. [16]. This method does not rely on morphological examination (i.e., the actual visualization of communicating veins) rather than on functional observation. After clamping of proper hepatic vein, the flow in the portal branch in the veno-occlusive area is evaluated with color Doppler. In the presence of an adequate outflow through the adjacent hepatic vein, portal flow remains hepatopetal thanks to intrahepatic venous anastomoses. In contrast, if the veno-occlusive area is congested, hepatic venous blood is regurgitated to the portal vein through the sinusoid, thus inverting the direction from hepatopetal to hepatofugal (Fig. 16.13). In this area, portal branches become draining veins, and the occluded area is supplied with arterial blood alone. This supply, especially in an engorged segment, could not be adequate and the parenchyma can become ischemic and eventually necrotic. Therefore it is advisable to resect that segment. We recommend checking the direction of the portal flow even in the presence of accessory or communicating veins. This technique allows operations such as the bisegmentectomy Sg7–8 without right inferior hepatic vein, which we first described in 2006; the bisegmentectomy Sg6–7 with RHV ligation, central resection with ligation of MHV sparing segment 5 and 4b.