Preoperative portal vein embolization

INTRODUCTION — Liver resection has evolved into an oncologically effective treatment for primary hepatic malignancies (eg, hepatocellular carcinoma, cholangiocarcinoma) and for the treatment of metastatic liver tumors (eg, colorectal carcinoma, neuroendocrine tumor). In many clinical situations, liver resection, when possible, provides the best survival outcomes. However, the limits of liver resection are determined by the probability of leaving behind a safe volume of functional liver that has an adequate vascular inflow, venous outflow, and biliary drainage.

The generally accepted safe minimum threshold for the future liver remnant (FLR) varies from 20 to 25 percent for a normal liver to >40 percent for a cirrhotic liver [1-3]. If the FLR volume is deemed unsafe or marginal, adjunctive or alternative methods to ensure an adequate postresection liver volume may include a staged resection, the associating liver partition and portal vein ligation for staged hepatectomy procedure, a hybrid procedure that combines liver resection with ablation techniques, and portal vein occlusion techniques such as preoperative portal vein ligation (PVL) or preoperative portal vein embolization (PVE).

Preoperative PVE is the elective obliteration of portal blood flow to a selected portion of the liver a few weeks prior to planned major liver resection. PVE initiates hypertrophy of the anticipated FLR. Preoperative PVE is a valuable adjunct to major liver resection, particularly for right-sided tumors, and it may allow a more extensive resection or staged bilateral resections [4-10]. The indications, contraindications, and techniques for PVE and a comparison of PVE with other techniques are reviewed here. (See "Overview of hepatic resection" and "Open hepatic resection techniques".)

LIVER VOLUME AND FUNCTIONAL ASSESSMENT — In evaluating the need for preoperative portal vein embolization (PVE), the standardized future liver remnant (FLR) needs to be determined. (See "Overview of hepatic resection", section on 'Preoperative imaging'.)

A clear understanding and application of appropriate estimation of the volume and function of the liver is an important aspect of patient selection. Many methods are in clinical use, and each has advantages and disadvantages. There is always a discrepancy between the measured or estimated volume and the actual weight, as demonstrated from graft weight measurements in living donor liver transplantation. The same applies to functional assessment as well.

The reasons for such discrepancies include:

●The assumption that the density of liver is 1.0 (ie, that the liver weight in grams = volume in mL). Although this is practically useful, the liver parenchyma is actually denser, and the conversion factor may be greater than 1.0.

●The assumption that liver parenchyma is homogenous. This is not true, as the proportion of the vasculobiliary components of the liver in relation to the hepatocellular mass is variable. Moreover, in patients with tumors, the size, number, and distribution will affect volumetry.

●In terms of function, the left lobe is marginally inferior to the right lobe even when corrected for volume [11].

Liver volumetry — Liver volume can be measured using manual or software-assisted methods using a number of imaging techniques such as computed tomography (CT), magnetic resonance (MR), or scintigraphy [3,12,13]. Alternatively, liver volume may be estimated using validated formulas that use a combination of age, sex, and anthropometric parameters such as height, weight, and body surface area [14]. Approximately 16 formulas are in use to derive estimated total liver volume (eTLV). The standardized functional liver remnant (sFLR) is the ratio between volumetrically measured FLR and the derived eTLV. Another method is the measured functional liver remnant (mFLR) in which both the remnant volume and the total liver volume are measured by CT volumetry. FLR in this method is usually expressed as a ratio of remnant volume to the total functional liver volume (total liver volume-tumor volume).

The chosen technique for measuring FLR depends upon the local availability of expertise and software. The results and principles of application are similar as long as the same technique is followed for both pre- and postembolization estimates. Thus, in the discussions below, we refer to FLR generically.

Functional liver studies — There is evidence from radionuclide studies that the increase in the functional capacity of the liver after preoperative PVE may be more than the increase in volumes that are measured using CT volumetry [15,16]. The increase in functional capacity can be measured by one of two radionuclide techniques: ^99mTc-labelled galactosyl-human serum albumin (^99mTc-GSA) or ^99mTc-labelled mebrofenin hepatobiliary scintigraphy with single-photon emission computed tomography (SPECT) [16]. The ^99mTc-GSA technique relies on the binding of the tracer to specific receptors expressed on functional hepatocytes, while the labeled mebrofenin with SPECT is based on the uptake and excretion kinetics of mebrofenin by the hepatocytes.

Other techniques used for functional assessment include indocyanine green retention at 15 minutes (ICG-15), the ¹³C-Methacetin breath test (LiMax test), and contrast-enhanced MR imaging using gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (gadolinium-EOB-DTPA) [16,17]. ICG-15 is an accepted parameter for functional capacity of liver. A level <10 percent is considered normal, although some labs quote a slightly higher value of up to 14 percent.

MECHANISM AND PHYSIOLOGY — Preoperative portal vein embolization (PVE) is the elective obliteration of portal blood flow to a selected portion of the liver a few weeks prior to planned major liver resection, with the intention of eliciting a hypertrophic physiologic response in the nonembolized portion. The purpose is to augment the volume and potentially the function of the future liver remnant (FLR) to a safe threshold and minimize the risk of post-hepatectomy liver failure and related complications such as sepsis, multiorgan failure, and mortality. Preoperative PVE usually takes the form of a right PVE, performed in preparation of a planned major right-sided resection such as an extended right hepatectomy (in a normal liver) or a right hepatectomy (in a cirrhotic liver). Very rarely, a left preoperative PVE may be indicated before a left-sided resection.

The sequela of selective portal vein occlusion, ipsilateral atrophy with contralateral hypertrophy, was first demonstrated in a rabbit model in 1920 [18]. Forty years later, portal vein ligation was used as a part of a two-staged hepatectomy [19]. However, the first PVE to facilitate safe liver resection was not reported until 1984 [20]. The effectiveness of PVE has been documented for a variety of different cancers and in patients with varying clinical profiles.

The increase in liver volume is due to clonal expansion of the number of hepatocytes (hyperplasia) and not by an increase in size of existing hepatocytes (hypertrophy). Preoperative PVE causes immediate redistribution of portal blood flow to the nonembolized liver segments, which returns toward baseline levels by 11 days [21]. The increased flow stimulates a regenerative response. The arterial flow to the nonembolized segments is reduced only marginally [22]. Key factors involved in the regenerative response include hepatocyte growth factor, epidermal growth factor, transforming growth factor alpha, insulin, noradrenaline, and cytokines such as interleukin-6 and tumor necrosis factor-alpha. The genes for some of these may be expressed as quickly as 30 minutes after hepatocyte injury [23].

At four to six weeks after PVE, the expected increase of standardized FLR ranges from 8 to 13 percent, although some studies have reported higher rates [5,24,25]. When expressed as a percentage augmentation from baseline FLR (ie, without PVE), this corresponds to a 40 to 62 percent increase [26]. (See 'Liver volume and functional assessment' above.)

In a retrospective study of 79 patients, four radiological features were identified as adverse factors for remnant hypertrophy after PVE [27]. These included: variant portal venous anatomy at the hilum (as against the standard bifurcation); presence of proximal small branches from the right anterior and posterior sectoral veins; presence of tumor thrombus in the right portal vein; and presence of transient parenchymal enhancement in the arterial phase. The absence of all these features predicted a favorable outcome (50 percent remnant hypertrophy), while the presence of any one of the features almost halved the remnant hypertrophy. Although promising, these features need validation from larger studies before they can be incorporated into selection algorithms.

The kinetics of liver regeneration following preoperative PVE are slower and of a different nature compared with those following liver resection. The rate of liver regeneration in normal liver peaks approximately two weeks after PVE at 12 to 21 cm³/day, and plateaus by three weeks, with a rate at one month of approximately 6 to 11 cm³/day [4]. However, studies have documented a continued growth of the liver up to one year post-PVE, with only about 25 percent of the total possible growth occurring in the first month, 50 percent by three months, and 75 percent by about eight months [28]. Factors that adversely affect regeneration after PVE include increasing age, male sex, a high initial FLR volume, diabetes, cholestasis (with or without biliary drainage), and parenchymal liver disease [5,29,30]. Some [31,32], but not all, studies [10,33,34] have found a detrimental effect of prolonged pre-PVE chemotherapy on the volume gain.

The growth rate after PVE as well as the degree of hypertrophy (DH), which is the difference between the FLR percentage before and after PVE, have been shown to be important predictors of post-hepatectomy liver failure [15,25,35]. The FLR percentage is the ratio of the FLR volume measured with computed tomography (CT) volumetry (numerator) and the functional liver volume (total liver volume-tumor volume) (denominator).

Where:

DH: degree of hypertrophy

TLV: total liver volume

FLR: future liver remnant

PVE: portal vein embolization

A DH >5 percent in normal liver and >10 percent in cirrhotic liver is considered a marker of adequate regenerative capacity and hence carries a reduced risk of postresection liver failure [10,36].

The kinetic growth rate (KGR) is the DH/week. A KGR >2 to 2.66 percent/week is considered to be a marker of a safe hypertrophic response that correlates with reduced rates of post-hepatectomy liver failure and major complications [25,35,37].

PATIENT SELECTION — Based upon observational studies demonstrating improved survival and reduced rates of postoperative liver failure following extended hepatic resections in those who have undergone portal vein embolization (PVE) [38-40], in the absence of severe liver dysfunction, we suggest an attempt at PVE in most patients with a marginal predicted future liver remnant (FLR; ie, <20 percent in patients with a normal liver, <30 percent for patients with nonalcoholic steatohepatitis, <40 percent in patients with cirrhosis) [41]. (See 'Functional liver remnant thresholds' below.)

However, the probability of successful liver enlargement decreases as the volume of FLR that is needed increases. In addition, as the severity of underlying liver disease increases, the likelihood that the liver will actually enlarge in response to PVE decreases [42]. Systemic disease such as diabetes may additionally limit liver hypertrophy and the success of the procedure. PVE appears to be most beneficial for patients with steatohepatitis with no metabolic dysfunction.

Benefits and risks — There are at least five potential benefits to preoperative PVE [4]:

●Postresection morbidity is diminished, as evidenced by minimal reductions in postresection liver function [43], and patients have fewer pulmonary complications and decreased length of intensive care unit and inpatient hospitalization.

●Patients who were initially unresectable because of insufficient remaining normal hepatic parenchyma can undergo resection for potential cure. In one study involving resection of extensive colorectal metastases, the rate of curative resection was increased from 46 to 79 percent following preoperative PVE [35].

●Subclinical disease or rapid progression may be detected prior to definitive surgery on postembolization imaging studies, thus preventing an unnecessary operation.

●The volume of remaining liver standardized to patient size (body surface area) predicts mortality after resection in cirrhotics, and the increase in functional liver volume achieved by preoperative PVE may decrease mortality of resection in these patients [4,43,44].

●The absence of compensatory hypertrophy in response to successful preoperative PVE indicates the patient is not suitable for major hepatic resection.

Risks specific to preoperative PVE include portal vein thrombosis, liver infarction and necrosis, and portal hypertension [45]. Some reports have shown accelerated tumor growth in the liver after PVE. However, tumor growth is not seen when all of the tumor-bearing areas of the liver are embolized [10].

Other risks related to percutaneous transhepatic access are similar to those of any percutaneous intervention, such as bleeding, infection, pseudoaneurysm, and arteriovenous fistula [45]. (See "Access-related complications of percutaneous access for diagnostic or interventional procedures".)

Functional liver remnant thresholds — Preoperative PVE is indicated when the FLR is deemed inadequate or unsafe and there is a reasonable prospect of an increase in the volume of FLR after preoperative PVE to an extent that would shift the expected FLR to a safe level that would permit liver resection. Although there is no universal consensus on what would be an ideal minimum FLR, most experts agree on the following broad practical guidelines for considering PVE [1,2]:

●In an otherwise normal liver (this is a rare situation in clinical practice):

•A standardized FLR of <20 percent, or

•FLR to body weight ratio of <0.5 percent (Truant criterion) [46]

●In the presence of significant steatosis/cholestasis/chemotherapy-associated steatohepatitis/chronic hepatitis:

•A standardized FLR of <30 percent [47], or

•FLR to body weight ratio of <0.8 percent

●In the presence of cirrhosis (Child A):

•A standardized FLR of <40 percent [48], or

•FLR to body weight ratio <1.4 percent [49], or

•FLR of <250 mL/m² (Shirabe criterion) [44]

However, a wide variation exists in the cutoffs used depending upon a number of other factors, such as age, diabetes, background liver disease, and individual surgeon experience. The evolution of liver resection has lowered these thresholds. In the early experience, the FLR threshold in a normal liver for PVE in the treatment of cholangiocarcinoma was 40 percent. In the early 2000s, the recommended minimum standardized FLR was lowered to 25 percent [12,50]. Further studies supported a threshold as low as 20 percent [10,51,52]. A consensus conference on resectability of liver metastasis in 2006 and another on multidisciplinary management of hepatocellular carcinoma in 2010 have endorsed the lower threshold of 20 percent minimum FLR in normal functioning livers [1,2].

Another point that should be kept in mind, particularly in a setting of hilar cholangiocarcinoma, is that PVE commits the surgeon to the ipsilateral resection. This often means an extended right hepatectomy. This places a significant constraint on the flexibility of operative decision making in patients with hilar cholangiocarcinoma who have had a PVE.

In cirrhotic livers, in addition to the Child-Pugh status, indocyanine green retention at 15 minutes (ICG-R15) is often factored into treatment algorithms. As an example, a protocol from China considers Child B or Child C status to be contraindications to major hepatic resection (>3 segments), and also an ICG-R15 of >20 percent [53]. But for patients with Child A cirrhosis, if the ICG-R15 is <20 percent, then a minimum FLR of 30 percent is required for a right hepatectomy and 35 percent for an extended right hepatectomy. Patients with FLRs below these thresholds are considered for PVE. Another major cancer center protocol in the United States uses a higher FLR threshold for hepatic resection in patients with Child A cirrhosis. If ICG-R15 is normal (<10 percent), then a 40 percent remnant is acceptable, while a 50 percent FLR is required if ICG-R15 is between 10 and 20 percent [36].

Contraindications — Contraindications to preoperative PVE include:

●Patient unfit for major resective surgery

•Significant cardiopulmonary comorbidities

•Poor performance status

•Child C status (most Child B patients as well)

•Severe portal hypertension

●Patient unfit for intervention (may be temporary)

•Uncorrectable coagulopathy

•Overt sepsis

•Renal failure needing dialysis

●Technical issues

•Lack of portal vein bifurcation

•Lack of intrahepatic portal vein (Abernathy malformation)

•Complete tumor thrombosis of the right portal vein

•Tumor thrombus extending into the FLR

•Tumor precluding safe transhepatic access (transsplenic access may be an alternative)

●Disease that is too extensive

•Extrahepatic disease precluding curative treatment

•Unlikely to achieve margin negative or R0 resection status in view of extensive disease in FLR

●Inadequate predicted FLR

•Baseline FLR so low that it is unlikely PVE would achieve an adequate FLR (this is usually an issue only in patients with chronic liver disease)

●Bilirubin >5 mg/dL (85 micromol/L) is a relative contraindication

●Biliary obstruction

●Ascites

TECHNIQUES

Procedure — Preoperative portal vein embolization (PVE) is usually performed under conscious sedation and local anesthesia, typically via a percutaneous approach in an interventional suite [54,55]. It can also be performed using an open approach, which is accomplished by cannulating the ileocolic vein at laparotomy, which requires a cooperative effort by the surgeon and interventional radiologist. Although an open approach was the original approach for PVE, it is rarely indicated, but may be used if laparotomy is indicated for another purpose, such as resection of a primary tumor in the colon.

Percutaneous PVE can be accomplished using a transhepatic or transjugular route. A transsplenic approach has also been reported for PVE in patients in whom multiple tumors precluded a safe trajectory of transhepatic puncture [56]. Direct transhepatic puncture of the portal vein under image guidance is the most commonly performed technique [5]. There are two approaches, the ipsilateral (same side as the tumor/intended resection) and contralateral approach, depending upon which portal vein is punctured [57]. The contralateral approach, first described in 1986, is technically easier but risks injury to the future liver remnant (FLR) [58]. The ipsilateral approach, initially described later in 1996, is more difficult but has become the favored approach [59]. (See 'Disease-specific outcomes' below.)

A variety of embolization agents have been used and include polyvinyl alcohol (PVA), ethiodized oil, absolute alcohol, fibrin, sodium tetradecyl sulfate foam, cyanoacrylate glue, gelatin, metallic spherical particles, and coils [57,60]. Absolute alcohol, although very effective, may cause severe inflammation and has fallen out of favor. Cyanoacrylate causes more effective and long-lasting occlusion compared with thrombin or gelfoam. PVA with coils also produces excellent occlusion with minimal reaction [4]. In general, PVE with small spherical particles is more effective (69 percent increase) than large, nonspherical particles (46 percent increase) [61]. The choice depends on the local expertise, availability, and cost.

Adjunctive techniques

Sequential TACE-PVE — There are potential benefits for performing a transarterial chemoembolization (TACE) before a PVE, especially in the setting of hepatocellular carcinoma (HCC), which is dependent on arterial blood supply. TACE will starve the tumor, and this may prevent interval progression of tumor post-PVE. The inflammatory response to TACE may also contribute to the regenerative response to PVE [22]. In addition, PVE alone might augment tumor blood flow due to the arterial buffer response. However, TACE can obliterate arterioportal shunts, which may cause an escape phenomenon blunting the effect of a PVE [62].

A number of studies have now supported the view that sequential TACE followed by PVE not only causes a greater and quicker increase in the FLR but also translates to a higher complete tumor necrosis rate and increase in recurrence-free and overall survival in HCC [7,63-66]. PVE can be performed one to two weeks after TACE [63]. For optimal effect in large HCCs that may derive additional blood supply from a segment IV artery or inferior phrenic artery, these vessels may also need to be embolized. (See "Surgical resection of hepatocellular carcinoma".)

Concomitant segmental arterial ligation and PVE — An intraoperative technique pioneered by the French called associating portal embolization and arterial ligation (APEAL) combines right PVE and ligation of a segmental branch of the right hepatic artery (anterior or posterior segmental) along with ligation of Glissonian pedicles to segment IVb (figure 1) [67]. This technique evolved in the context of a staged procedure for multiple colorectal liver metastases. The first stage consisted of resection of the colorectal primary, clearance of any tumor in the FLR by nonanatomic resection or ablation, and the APEAL procedure. The second stage that was performed after a median of 45 days consisted of an extended right hepatectomy. Median FLR hypertrophy rates of 104 percent at seven days and 134 percent at 30 days have been reported [67]. The attractive aspects of the APEAL procedure are the augmented FLR hypertrophy over and above those observed after conventional PVE and the reduced morbidity and mortality compared with the associating liver partition and portal vein ligation for staged hepatectomy procedure [67]. (See 'PVE versus the ALPPS procedure' below.)

Other techniques

●Sequential PVE and hepatic vein embolization (PVE-HVE) – Hepatic vein embolization (HVE) after PVE may further increase FLR hypertrophy. It may also improve the development of venous collaterals in the FLR [68,69]. PVE-HVE is usually used as a salvage option when PVE alone fails to achieve adequate FLR hypertrophy. Some centers elect to combine the procedures simultaneously in a single setting and have reported good oncological outcomes [70]. When PVE is combined with HVE, the technique has been called liver venous deprivation.

●PVE with stem cell transplantation – Bone marrow-derived progenitor cells (CD 133⁺ cells) play a role in repopulation and regeneration of the liver [71]. Infusing these cells into the FLR soon after PVE has been shown to result in a greater and quicker FLR hypertrophy without increasing the complication rate [72-74].

●Branched chain amino acid supplementation – There is evidence from a small randomized trial that branched chain amino acid supplementation with PVE increases FLR function [75]. Although the evidence is limited, this may be a practical, simple, and safe adjunct in patients undergoing PVE.

Experimental techniques — Experimental variants of PVE, none of which has been clinically tested, include Madoff's transarterial PVE, Smits's trans-sinusoidal PVE, and Lainas's reversible PVE [8,76].

●Madoff's transarterial PVE involves a slow microcatheter infusion of a mixture of iodized oil and absolute alcohol (3:1 ratio) as the embolic medium via the lobar branches of the hepatic artery, which is supposed to reach the portal system through peribiliary plexus. The authors claim a twofold advantage in hypertrophy volumes based on a small series in pigs [77].

●Smits's trans-sinusoidal PVE involves injecting a low-viscosity medium into the target hepatic vein while wedged. This relies on reflux of the medium into the portal vein via the sinusoids [78].

●Reversible PVE, as the name implies, involves employing a specific agent such as powdered gelatin sponge, the embolic effect of which is temporary. The small population that might benefit from this technical variant are those who do not undergo resection for various reasons post-PVE [79]. Using this technique, animal studies have widely variable results as to the degree of FLR hypertrophy.

AREAS OF CONTROVERSY

Timing of resection after PVE — The timing of hepatic resection following preoperative portal vein embolization (PVE) is a balance between the risk of tumor progression and the benefit of more gain in liver volume with passing time. In general, hepatic surgeons wait for three to six weeks post-PVE before undertaking resection [10,80,81].

Volumetric assessment of the liver volume using computed tomography or magnetic resonance imaging is performed before PVE and repeated four weeks after PVE to assess the extent of liver hypertrophy prior to undertaking surgery [45,50]. In patients found to have an insufficient future liver remnant (FLR), the study scan can be repeated after a further three to four weeks to assess further for hypertrophy. (See 'Liver volume and functional assessment' above.)

The increase in liver function outpaces the increase in volume, which always lags behind. Although the rate of growth peaks early and tends to plateau at approximately three weeks, the volume gain at one month is only approximately 25 percent of the total possible volume gain. The liver continues to increase in volume as long as one year after preoperative PVE. (See 'Mechanism and physiology' above.)

Extended PVE — Right PVE is often performed when an intended right trisectionectomy is planned (figure 2). Standard right PVE will result in segment IV hypertrophy (as well as hypertrophy of the left lateral segments) (figure 1). However, the response in the left lateral segments is important since segment IV would be resected and hence will not contribute to the FLR. Embolization of the segment IV portal vein branch along with the right portal vein (ie, extended embolization) was first reported in 2000 as right trisegment embolization [82]. If an ipsilateral approach to PVE is used, then segment IV embolization should be done first before right PVE; on the other hand, with a contralateral approach, the right PVE is done first before segment IV embolization. It is also essential to embolize the segment IV branches if segment IV is tumor bearing. Rarely, a left trisegment embolization, which includes embolization of the left main portal vein and the right anterior portal vein, may be indicated if the right posterior segment volume is deemed inadequate [83].

There is evidence to suggest that with extended embolization, the rate and absolute increase in volume of segments II and III would be greater [82,84]. One study found an increase in FLR with extended embolization compared with standard right PVE (50 versus 31 percent), while another study similarly found double the response rate with extended embolization (percent FLR hypertrophy: 54 percent versus 26 percent) [82,84]. However, not all studies have found an advantage of extended embolization, which is technically more demanding and risks injury to the FLR [85]. No strong recommendations can be made, and the choice will depend upon available local expertise.

Potential acceleration of tumor growth — Approximately one sixth to one third of patients who undergo PVE for colorectal liver metastasis will suffer disease progression, making them unresectable [86]. There is a concern that at least part of this may be an effect of PVE itself. However, tumor growth may not always be the case, as tumor regression after PVE is also known to occur when PVE is preceded by chemotherapy [34,87].

PVE appears to accelerate tumor growth [88-92]. One study quantified tumor growth rate after PVE at 0.36 cc/day, which was higher than the 0.05 cc/day found in controls [91]. In colorectal metastasis, tumor growth tends to be similar in the embolized and nonembolized lobes [34,86,87]. Tumor growth rate acceleration in hepatocellular carcinoma is higher at 2.65-fold compared with cholangiocarcinoma at 1.16-fold [90]. A long gap between the last cycle of chemotherapy and PVE, as well as an initial nonresponse to chemotherapy, increases the risk of accelerated tumor growth [34,92].

Ongoing neoadjuvant chemotherapy post-PVE — For patients undergoing neoadjuvant chemotherapy prior to hepatic resection, we recommend continuing the chemotherapy following PVE.

Neoadjuvant chemotherapy is a commonly used strategy in managing initially unresectable or borderline resectable colorectal cancer liver metastasis, but it may also be applicable in selected patients with hilar cholangiocarcinoma. (See "Potentially resectable colorectal cancer liver metastases: Integration of surgery and chemotherapy", section on 'Patients undergoing preoperative portal vein embolization'.)

Whether neoadjuvant chemotherapy should be continued or stopped after PVE is controversial. Stopping chemotherapy after PVE while awaiting hypertrophy for four to six weeks may increase the risk of tumor progression. However, there is some concern that continuing chemotherapy may retard the hypertrophic response to PVE; such an effect has been observed with liver regeneration after liver resection [93-95]. However, this phenomenon may be less pronounced with PVE, and the evidence in fact supports an overall benefit from chemotherapy post-PVE. In a small study of 15 patients, hypertrophy did occur post-PVE despite the continuation of chemotherapy, albeit to a slightly lesser extent [96]. Larger studies have similarly shown no negative effect on the hypertrophic response [24,33,97]. More importantly, in a study of 208 lesions in 64 patients with colorectal liver metastasis, there was a significantly lower rate of tumor progression in patients who received post-PVE chemotherapy compared with those who did not, and continuing chemotherapy post-PVE was independently associated with prolonged long-term survival [86].

Postresection regeneration — There has been a concern that PVE may exhaust the regenerative capacity of the liver, and this may lead to inadequate regeneration after liver resection. However, PVE does not appear to adversely affect regeneration of liver after resection. In a small study, the liver regenerated to the same extent (ie, to 80 percent of its original volume) in patients who had PVE as with those who did not have a PVE [40]. (See "Overview of hepatic resection", section on 'Liver function and regeneration after resection'.)

DISEASE-SPECIFIC OUTCOMES — Preoperative portal vein embolization (PVE) allows potentially curative liver resection in patients with an inadequate future liver remnant (FLR). Prior to PVE, all of these patients would otherwise have been candidates for only palliative treatment. In general, approximately 70 to 80 percent of patients who undergo PVE ultimately undergo resective surgery [26].

Hepatocellular carcinoma — Resection for hepatocellular carcinoma (HCC) differs from other cancers because HCC often occurs in a setting of chronic liver disease. PVE is indicated only for well-compensated patients (Child A) with an acceptable hepatic functional reserve (indocyanine green retention at 15 minutes <20 percent, FLR 40 to 50 percent) and no other comorbidities that would preclude a major liver resection. (See "Overview of treatment approaches for hepatocellular carcinoma" and "Surgical resection of hepatocellular carcinoma".)

PVE improves resectability rates in HCC, and five-year survival rates after major liver resection post-PVE are at least equivalent and, in many studies, better than in those who did not need PVE [98-102]. As discussed above, sequential transarterial chemoembolization portal vein embolization has volumetric and oncologic advantages in HCC. (See 'Sequential TACE-PVE' above.)

Biliary cancer — Although technical success rates are high and procedure-related morbidity is low, in the largest review of patients with biliary cancer, approximately 25 percent of patients were unable to undergo liver resection [83]. This dropout rate was higher for gallbladder cancer compared with hilar cholangiocarcinoma (43 versus 17 percent). The five-year survival of resected patients was 39 percent for cholangiocarcinoma and 23 percent for gallbladder cancer, which is commendable given that without PVE all these patients would have otherwise been unresectable [83]. (See "Surgical resection of localized cholangiocarcinoma".)

Colorectal cancer liver metastasis — PVE for colorectal cancer liver metastases poses some unique challenges. These patients have often undergone neoadjuvant chemotherapy, and they may have some metastasis in the FLR. (See "Potentially resectable colorectal cancer liver metastases: Integration of surgery and chemotherapy", section on 'Patients undergoing preoperative portal vein embolization'.)

Disease burden in the FLR has to be managed before PVE can be applied. In a cohort of patients needing an extended right hepatectomy and presenting with primarily an inadequate FLR alone, PVE improved the rate of resectability from approximately 46 to 79 percent [35]. The perioperative surgical morbidity and mortality, and long-term survival of patients undergoing major hepatic resection after PVE, are reported to be similar to those of patients who had an adequate FLR at presentation and were not in need of PVE [35,103]. A systematic review identified six studies comparing outcomes of patients undergoing major liver resection for colorectal liver metastases and found no significant differences in postoperative hepatic recurrence for those undergoing liver resection with or without PVE [104].

Comparison with other techniques

PVE versus portal vein ligation — Portal vein ligation (PVL), performed by laparotomy or by laparoscopy, is an alternative to PVE [105]. Animal studies suggest that hepatocyte regeneration after PVL is more pronounced than after PVE [106]. A meta-analysis of seven clinical studies involving 218 patients found that there was no significant difference when comparing PVE with PVL in terms of FLR hypertrophy (39 percent with PVE versus 27 percent with PVL) [107]. The morbidity, mortality, and disease progression rates were also similar.

The current indications for PVL for augmenting FLR are:

●Interventional radiological facilities for PVE are not available

●The patient requires laparoscopy for initial staging

●The liver remnant is judged at laparotomy to be marginal

●The patient will be undergoing a two-staged resection, wherein the first stage involves clearing the FLR of tumor

●PVL is being performed as part of the associating liver partition and portal vein ligation for staged hepatectomy (ALPPS) procedure

PVE versus transarterial radioembolization — Transarterial radioembolization with 90-Yttrium(⁹⁰Y)-labeled microspheres, also known as selective internal radiotherapy (SIRT), is also known to induce FLR hypertrophy and has the added advantage of preventing tumor progression. However, the magnitude of hypertrophic response is inferior to PVE (29 versus 61 percent) [108].

PVE versus the ALPPS procedure — The ALPPS procedure is an alternative to preoperative PVE [26,109-116]. ALPPS is a short-interval, two-staged liver resection [117]. The first stage involves an open right PVL and in situ parenchymal transection in the right trisectionectomy plane [117]. The second stage is performed one to two weeks later and involves a right trisectionectomy. ALPPS produces an accelerated hypertrophy of the functional liver remnant (FLR) in a much shorter time interval than PVE. In a retrospective review, the extrapolated liver growth rate was 11 times higher for ALPPS (34.8 mL/day versus 3 mL/day) [113]. In a case series of patients undergoing ALPPS, the average hypertrophy rate was 85 percent [114].

ALPPS is performed only in specialist high-volume centers. In most centers, for patients with an inadequate FLR, preoperative PVE remains the preferred technique [26]. This view may change or be redefined in future as the expertise for ALPPS becomes more widely available. Typical drivers for ALPPS are the improved hypertrophy rates and a higher proportion of patients completing the two-staged treatment; however, the patient is exposed to additional operative risk [115,116]. Salvage ALPPS may have a role in the treatment of patients who do not show an adequate hypertrophy response to preoperative PVE. The LIGRO trial, a multicenter Scandinavian study, is the only randomized trial that has compared ALPPS head-to-head with a two-stage hepatectomy (TSH) where PVE was a component of TSH [118]. The resection rate for the ALPPS group was significantly higher compared with TSH (92 versus 57 percent) among 97 patients with colorectal liver metastasis. Among the 13 patients in the TSH arm who failed to achieve an FLR of 30 percent after PVE, 12 were successfully treated with ALPPS. When indicated, ALPPS should be performed at hepatobiliary centers of excellence that have experience with ALPPS.

MORBIDITY AND MORTALITY — Preoperative portal vein embolization (PVE) is overall well tolerated. In a systematic review that included 29 studies reporting on complications after preoperative PVE, the mortality rate was 0.1 percent [80]. The two deaths that followed PVE were due to late septic shock from cholangiolitic abscess, which may be attributed to PVE, and late pulmonary embolism.

Major complications are uncommon (2 to 3 percent) and include portal vein thrombosis, embolization of nontarget vessels, infection, bile leak, hemobilia, pseudoaneurysm, arteriovenous fistula, arterioportal shunt, liver hematoma, and pneumothorax [80,119-121]. Variceal bleeding from rise in portal pressure has also been reported [119].

Minor complications, such as fever and ill-defined abdominal discomfort with or without accompanying elevation of transaminases, are fairly common and self-limited, occurring in approximately 20 to 30 percent of patients. Transaminases peak at less than three times the normal values by post-PVE day 3 and return to baseline by 10 days. There may be a mild increase in bilirubin, but albumin and prothrombin time are not usually affected [4]. Ileus and vomiting are uncommon (1 to 2 percent) [80].

Technical failures occur in only 0.4 percent. The dropout rate after PVE is approximately 25 percent for various reasons such as disease progression and inadequate hypertrophy, which occurs in approximately 3 percent [80].

SUMMARY AND RECOMMENDATIONS

●Preoperative portal vein embolization – Preoperative portal vein embolization (PVE) is the elective obliteration of portal blood flow to a selected portion of the liver a few weeks prior to planned major liver resection. PVE elicits a hypertrophic physiologic response in the nonembolized portion augmenting the volume and potentially the function of the future liver remnant (FLR) to a safe threshold to permit potentially curative liver resection. (See 'Introduction' above and 'Mechanism and physiology' above.)

●Physiology – Preoperative PVE causes a modest increase in the standardized FLR in the range of 8 to 13 percent over a relatively short time of four to six weeks. The absolute FLR percentage increase is between 40 and 60 percent of the baseline FLR volume. (See 'Mechanism and physiology' above.)

●Patient selection – For patients with a marginal predicted FLR that precludes an otherwise potentially curative hepatectomy, we suggest preoperative PVE, rather than no hepatic resection or an alternative procedure (eg, portal vein ligation [PVL], associating liver partition and portal vein ligation for staged hepatectomy [ALPPS]), prior to hepatic resection provided there are no contraindications to the PVE (Grade 2C). Preoperative PVE is well tolerated and has a more favorable safety profile compared with available alternative techniques. (See 'Comparison with other techniques' above and 'Disease-specific outcomes' above and 'Contraindications' above.)

Preoperative portal vein embolization is indicated for a predicted future liver remnant that is (see 'Patient selection' above):

•<20 percent in a normal liver

•<30 percent in the context of cholestasis/hepatitis/chemotherapy

•<40 percent in Child A cirrhosis

●Timing of hepatic resection – Although the timing varies, hepatic surgeons generally wait three to six weeks post-PVE before undertaking hepatic resection. Approximately 70 to 80 percent of patients who undergo preoperative PVE eventually undergo successful liver resection. In properly selected patients, resectability rates are increased and the incidence of post-hepatectomy liver failure is decreased. (See 'Timing of resection after PVE' above and 'Disease-specific outcomes' above.)

●Techniques – Preoperative PVE is typically performed via a percutaneous transhepatic approach in an interventional suite under conscious sedation with local anesthesia.

•When an extended right hepatectomy is planned, then extended embolization to include segment IV branches of the portal vein should be performed if technical expertise is available. An open surgical approach may be used if laparotomy is indicated for another purpose. (See 'Techniques' above and 'Extended PVE' above.)

•Adjunctive techniques to PVE include transarterial chemoembolization (TACE), concomitant segmental arterial ligation, hepatic vein embolization, and stem cell transplantation. For patients with hepatocellular carcinoma, sequential TACE-PVE has volumetric as well as survival benefits. (See 'Adjunctive techniques' above.)