Techniques of hepatic resection
History
Liver surgery has been described for centuries in literature in relationship to patients being treated for stab wounds and other injuries to the liver. However it wasn’t until development in general anesthesia and antibiotics that formal liver resections became more prevalent. In 1886, Dr. Luis performed the first liver surgery but the patient died 6 hours later due to bleeding. The first successful liver resection is attributed to Dr. Langenbuch in 1888, although the patient was reoperated on for bleeding (1). These complications did not stop surgeons from studying the anatomy to find successful ways to resect liver tumors. In 1897, Dr. Cantlie’s description of the liver established further understanding of liver anatomy, leading to better control of bleeding during surgery. Kousnetzoff and Pensky in 1896 described the suture fracture technique that allows compression and ligation of vasculature while transecting the parenchyma, a technique that is still used today with some variations (1). One of the seminal contributions to liver resections came from Dr. Pringle, who in 1908 described the technique of compressing portal inflow to decrease liver bleeding. Further improvement in operative exposure using subcostal incisions with retraction has also lead to a decrease in operative morbidity and mortality (1). In the last 60 years, technological advances have led to a rapid development in various techniques in liver resection. In this review article, we discuss specifically techniques in liver resection for metastatic colorectal cancer to the liver. We find that there is not just one specific way that works the best. Tumors in different lobes of the liver will require a different approach with the specific technique tailored to each patient.
Anatomy
Surgery of the liver is based largely on the anatomic description of functional segments, which in turn is based on the organ’s blood supply via the hepatic artery and portal vein, its venous drainage via the hepatic veins, and finally, its biliary drainage. This division of the liver into eight functional segments is the most widely-accepted anatomic definition used in the context of hepatic resections (2-4). Major hepatic resections may be safely accomplished by adequately comprehending this internal segmental anatomy and its relationship to the major vascular structures (Figure 1).
The anatomic right and left lobes of the liver are divided by the ligamentum teres and umbilical fissure, where the main vascular and biliary structures to the functional left liver run. However, the true functional division of the right and left liver is divided by the middle hepatic vein. This can be demarcated by a plane extending from the left side of the gallbladder fossa anteriorly, to the left side of the inferior vena cava posteriorly (known as Cantlie’s line). The right and left liver are further subdivided into segments which follow the distribution of the portal triad structures. The right, middle, and left hepatic veins drain into the vena cava and run within the corresponding scissurae.
The left liver is divided by the falciform ligament into a medial and lateral segment. The left lateral segment is divided into a superior (segment II) and inferior segment (segment III) by the left portal vein. The left medial segment (segment IV) is also divided into a superior portion (IVa) and an inferior portion (IVb). These divisions correlate to branches from the portal vein. The right liver is divided into an anterior (segments 5, 8) and posterior segments (segments 6, 7) by the right hepatic vein. These segments are further subdivided into inferior and superior segments by the right portal vein. Thus, there are four segments that comprise the right liver: anteroinferior (medial, segment V), posteroinferior (lateral, segment VI), posterosuperior (lateral, segment VII), and anterosuperior (medial, segment VIII). The caudate lobe (segment I) is posterior and inferior in relationship to the rest of the liver, and lies over the inferior vena cava. It receives portal irrigation from both right and left branches and drains directly into the vena cava.
The terminology of major hepatic resections arises from the segmental anatomic description above (Figures 1, 2). Right hepatectomy (or hemihepatectomy) involves resection of segments V-VIII, whereas left hepatectomy involves resection of segments (II-IV). Right lobectomy (also known as extended right hepatectomy, or right trisegmentectomy) involves resection of all segments lateral to the umbilical fissure (IV-VIII, and sometimes I), whereas extended left hepatectomy (or left trisegmentectomy includes resection of all liver medial to the umbilical fissure and a portion of the right liver (segments II-IV and segments V and VIII). Left lobectomy (also known as left lateral segmentectomy) involves resection of all liver medial to the umbilical fissure only (segments II and III) (1,5).
Types of Major Resections
An important decision in any liver resection is choosing the amount of parenchyma to be removed. Anatomic resections usually involve 2 or more hepatic segments, while non-anatomic resection involves resection of the metastases with a margin of uninvolved tissue (segmentectomy). This decision regarding extent of resection becomes especially relevant in the setting of post preoperative chemotherapy in colorectal metastasis, where an attempt to maximize the remnant liver volume is made. While preoperative therapy allows more patients to be considered resectable, it can compromise hepatic function and increase the risk of postoperative liver failure (6). Thus, the choice to perform a non-anatomic, or wedge resection should consider key factors such as preoperative chemotherapy, pre-existing liver disease, tumor burden, risk of recurrence, and whether or not outcome will be affected by the extent of resection (7). The greater parenchymal-sparing surgery afforded by a non-anatomical resection may prove to be beneficial especially in the setting of intrahepatic recurrent disease, which occurs in up to 50% of cases, where local minimally-invasive ablative therapies may be more amenable.
A small series of patients who underwent initial partial hepatic resection and recurred thereafter was reported by van der Pool and colleagues. They demonstrated that repeat treatment for recurrence of intrahepatic disease with local therapies (which included repeat non-anatomic resection, radiofrequency ablation, or stereotactic radiation) can be performed safely and with good median overall survival (37 months) and an overall 5-year survival rate of 35% in their series (8). A recent Dutch retrospective study compared the difference in morbidity and mortality and the patterns of recurrence and survival in 201 patients with colorectal liver metastases treated initially with anatomic versus nonanatomic liver resection. The trial found that nonanatomic resection was typically performed for significant smaller metastases, and this group received significantly less blood transfusions and had a shorter hospital stay. The groups did not differ with regards to morbidity, mortality, recurrence rate, or survival according to resection type (9). In similar fashion, multiple previous studies comparing anatomic vs. non-anatomic resection for colorectal liver metastases have not demonstrated any significant differences with regards to survival, margin status, or patterns of recurrence (10-12).
Vascular Control
Blood loss is among the most important variables influencing postoperative outcome from hepatic resection (13). In order to perform liver resections safely and to minimize blood loss and need for blood transfusions, it is essential to be familiar with different hepatic vascular occlusion techniques available. The application of each individual technique should be based upon the type of resection to be performed, tumor size and location, and preoperative liver function. More importantly, the different methods of vascular control each have distinct physiologic and hemodynamic effects systemically and within the liver itself, and thus the choice of which method to use should be determined by the patient’s ability to tolerate it. The array of vascular occlusion techniques ranges from Pringle’s maneuver (portal triad clamping) to total hepatic vascular exclusion, including inflow occlusion (selective or total), hemi-hepatic clamping, and ischemic pre-conditioning. These methods can also vary with regards to timing and frequency (intermittent vs. continuous) (14).
Inflow occlusion by hepatic pedicle clamping has been shown to reduce blood loss during liver resection (15). This is a consistent method of vascular control, which is not technically very difficult to perform. While it addresses the portal vein and hepatic artery, it does not address backbleeding from the hepatic veins. The Pringle maneuver can be performed continuously or intermittently and is usually well tolerated by the liver. When performed intermittently, the portal triad is typically clamped for 10 minutes and then unclamped for 3 minutes (the clamping on and off can vary). This allows for a longer total occlusion time of up to 2 hours in the normal liver, which can be useful for more prolonged complex liver resections, as demonstrated in previous studies (16). The increased blood loss during the periods of unclamping can be a challenge; however, the total blood loss or transfusion requirements does not differ between the intermittent and continuous techniques (17).
A potential consequence of the intermittent technique is hepatocyte injury from a sequence of ischemia-reperfusion periods. However, a prospective, randomized study by Clavien, et al. demonstrated that a 10 minute sequence of ischemia and reperfusion preceding a longer 30 minute period of continuous vascular occlusion was a protective strategy in humans. In their study, these findings were more effective for younger patients requiring a prolonged period of inflow occlusion (18). This strategy of ischemic preconditioning when compared with intermittent Pringle did not show any significant difference in terms of blood loss in a recent meta-analysis (19).
The continuous Pringle maneuver allows for shorter total occlusion time, and has the advantage of avoiding interruption of the parenchymal transaction (20). Belghiti and colleagues nevertheless demonstrated that this does not necessarily translate into shorter overall operative time (21). Both the continuous and intermittent methods should be used for shorter time periods in the setting of chronically diseased livers or patients that have undergone preoperative chemotherapy. In the setting of chronic liver disease, the intermittent method has been shown to be better tolerated (22).
Total hepatic vascular exclusion is another method of reducing blood loss during liver resection by occluding the inflow and outflow. This technique mitigates the risk of retrograde hepatic vein bleeding and can decrease the risk of air embolism. Hepatic vascular exclusion is more technically difficult than pedicle clamping alone, as it requires complete mobilization of the liver and appropriate exposure of the inferior vena cava. This method may be performed by clamping the portal triad in addition to clamping the infrahepatic and suprahepatic vena cava, or more selectively by clamping the hepatic veins extraparenchymally and preserving caval flow. One of the major challenges of total hepatic vascular occlusion is the hemodynamic effects it induces, which may be poorly tolerated in up to 15% of patients (14). There is a 40-60% decrease in cardiac output and blood pressure, with the resulting compensatory mechanisms of tachycardia and increased systemic vascular resistance (23). It is associated with an increased risk of postoperative complications, increased operative time, and lacks significant benefit over portal triad clamping alone with regards to blood loss, transfusion requirements, and liver failure (14,24,25).
Another technique important in decreasing blood loss and operative time involves intrahepatic pedicle ligation. Ligation of the right, left or smaller branches of the portal vasculature supplying the portion of liver being resected is an important step in liver resection. The previous mentioned techniques of vascular control are important for controlling back bleeding from the adjacent segments of liver during transection. Understanding the anatomy of the portal vessels permits a safe approach to pedicle ligation. Portal triad consists of common hepatic duct, portal vein and hepatic artery (Figure 1). The triad is encased by Glisson’s capsule. As the portal triad enters the hilum of the liver, it splits into the right and left portal pedicles. The right further splits into the anterior and posterior branches. The left travels in the umbilical fissure and gives branches to segments II, III, IV. For right hepatectomy, a hepatotomy is made in segment IV in the gallbladder fossa and a second one made inferior to the main right portal branch near the caudate lobe. A blunt right angle or renal pedicle clamp is then passed superior from the hepatotomy in segment IV through liver parenchyma and then exiting via the inferior hepatotomy. A vascular clamp is used to compress this tissue and right portal pedicle allowing for demarcation of the right lobe. Once this is confirmed, a vascular stapler is used to transect the pedicle.
For a left hepatectomy, the hilar plate is elevated and the left portal pedicle is identified in the umbilical fissure. A hepatotomy is made at the level of lowering the hilar plate and a second hepatotomy in the back of segment II. The same clamp should be used to come around this pedicle with subsequent vascular clamping to check for demarcation and then a vascular stapler to transect the left portal pedicle. This technique used properly can decrease blood loss, decrease risk to injury of the hilum, and shortens operative time. Use of intraoperative US during pedicle ligation decreases injury to nearby vasculature. Pedicle ligation can be used in select cases needing segmentectomy. However, this maneuver should not be used for centrally located tumors because obtaining surgical margins will not be possible. For patients that cannot undergo pedicle ligation, the standard technique of isolating the hepatic artery, portal vein separately should be performed (extrahepatic ligation).
Parenchymal Transection
As the number of liver resections have increased over the past 20 years, so too has the armamentarium of surgical devices available to facilitate the different aspects of liver surgery such as vascular control, hemostasis, and parenchymal transection. This growing variety of tools has been especially represented in the field of parenchymal transection. The methods range from basic finger or clamp-fracturing the tissue, to devices based on more complex technology, such as ultrasonic or radiofrequency energy, water jet and tissue-sealing devices, and surgical staplers. These strategies are all aimed at reducing blood loss and transfusion requirements, and the increased postoperative complications associated with each. Additionally, there are other important factors to be considered when choosing a particular method, such as operative time, availability and ease of use, extent of hepatic injury affected, and cost. The use of one tool over the other will also vary according to the type of resection, and different techniques can be more advantageous in one setting than another. It is important to be familiar with many strategies and be able to apply them in the most appropriate setting. We discuss the most widely used methods at present and review the existing randomized data comparing them.
Crushing Technique
The most basic strategy involves crushing the liver parenchyma between the surgeon’s fingers in order to expose and isolate small vessels and biliary radicals, which can then be divided. This same fundamental technique can be further enhanced by employing basic surgical clamps (so-called “crush-clamp” technique) to crush the hepatic parenchyma, as shown (Figure 3). The crush-clamp method usually affords superior control when transecting the parenchyma as compared to the finger fracture method. Once the parenchyma is crushed, the exposed vessels and bile ducts can be divided. The latter can be achieved by silk suture ligation, bipolar electrocautery, vessel sealing devices, or vascular clips. Intermittent inflow occlusion with the Pringle maneuver is typically used during the transection and coagulation (Bovie cautery or argon beam coagulation) is applied to the remnant liver parenchyma during the periods of reperfusion for hemostasis. This technique is simple, quick, efficient, easy to learn and perform, and cost-effective. The crush-clamp strategy has served as the point of reference for all other hepatic parenchymal transection techniques. A series of randomized controlled trials and subsequent meta-analyses discussed below have analyzed and compared this method with newer ones.
A trial from Switzerland randomized 100 patients without cirrhosis or cholestasis to undergo liver resection using one of four methods: crush-clamp, ultrasonic dissector, water jet, or dissecting sealer (26). The patients randomized to the crush-clamp technique all underwent major hepatectomy with vascular inflow occlusion using a continuous Pringle maneuver, as opposed to the other groups in which routine Pringle maneuver was not used. The crush-clamp technique was associated with a shorter resection time, less blood loss, lower frequency of blood transfusion, and proved to be the most effective method. A subsequent German meta-analysis by Rahbari and colleagues analyzed seven randomized controlled trials with greater than 500 patients and found no clinically important benefit of an alternative transection method in terms of blood loss, parenchymal injury, transection time, and hospital stay (27). In similar fashion, a 2009 Cochrane review of randomized data failed to show any significant differences with regards to mortality, morbidity, markers of liver parenchymal injury, or ICU/hospital length of stay when comparing crush-clamp to alternative methods (28). The review did show crush-clamp to be faster and less expensive as well. Finally, the CRUNSH trial is a newly-designed prospective, randomized controlled trial comparing the efficacy of the crush-clamp technique versus use of a vascular stapler for parenchymal transection (29).
Ultrasonic Dissection
The Cavitron Ultrasonic Surgical Aspirator (CUSA, Tyco Healthcare, Mansfield, MA, USA) (Figure 4), combines ultrasonic energy with aspiration in order to divide the liver parenchyma and thus skeletonize small parenchymal vessels and biliary structures greater than 2mm, which are then divided according to preference. CUSA is able to dissect tissue but does not offer coagulation or hemostasis. Among CUSA’s benefits, it provides a very well-defined transection plane, which is useful in situations of close proximity between tumors and major vascular structures. Also, it can be used in cirrhotic as well as non-cirrhotic livers, and is associated with a low blood loss and low risk of bile leak (30). Transection time using CUSA is generally slower than conventional methods. Nonrandomized studies have shown decreased blood loss, morbidity, and mortality using CUSA, however, larger randomized trials have not shown this benefit over the traditional crush-clamp method (31).
The Harmonic Scalpel (Ethicon Endo-Surgery, Cincinnati, OH, USA) uses a similar principle of ultrasound energy applied to vibrating ultrasonic shears to seal and divide blood vessels up to 3mm in diameter. The vibration of the blades at 55,500 times per second simultaneously cuts and coagulates tissues by causing denaturization of proteins, rather than heat, as with conventional electrocautery. This allows for a more precise transection plane and reduces lateral thermal damage as well. The Harmonic Scalpel finds its best application during laparoscopic liver resections. In a nonrandomized study by Kim, et al. use of the Harmonic Scalpel was associated with decreased operative time and a trend toward decreased blood loss and transfusion requirement. However, it was also associated with a significant increase in the incidence of postoperative bile leaks (