Neuroendocrine liver metastases derive their blood supply from the hepatic arteries, while normal hepatocytes derive most of their blood supply from the portal vein. Transarterial therapies can thus target NET metastases with relative sparing of normal liver. Transarterial therapies play a central role in the management of NET liver metastases since the majority of patients are not amenable to surgical resection and/or ablation at the time of diagnosis.5 Furthermore, because NET liver metastases invariably recur after resection, there is often a role for transarterial therapies even after successful surgery and ablation.6,24 These therapies include transarterial embolization (TAE), transarterial chemoembolization (TACE), and selective intra-arterial radiation therapy (SIRT or radioembolization). Each of these therapies are discussed in detail in the following sections.
Clinical and Technical Considerations
Similar to that discussed previously with thermal ablation, planning for TAE begins with clinical consultation and review of cross-sectional imaging. Clinical consultation is important for identifying comorbidities and for managing patient expectations. Routine laboratory and coagulation testing is performed. For transarterial procedures, a goal international normalized ratio (INR) of <1.5 and platelets >50,000 is recommended. Clopidogrel is held for 5 days prior to the procedure and low-molecular-weight heparin (LMWH) dose held on the day of procedure. Aspirin therapy may be continued depending on the clinical scenario.12
Transarterial therapies are typically performed under moderate sedation and with local anesthesia at the arteriotomy site. Broad-spectrum antibiotics are recommended prior to the procedure.11 In addition, octreotide therapy is typically initiated before and during embolotherapy to prevent potential adverse effects (i.e., malignant hypertension) of preformed hormones that may be released with tumor necrosis. The arterial puncture is made in the femoral artery distal to the inguinal ligament and overlying the femoral head. Puncture at the level of the femoral head is crucial to allow for adequate compression and hemostasis at the conclusion of the procedure.
Prior to embolization, angiography of the celiac and superior mesenteric arteries is performed to identify hepatic arterial anatomy. In one review of 600 patients, only 61% of patients had traditional hepatic arterial anatomy with left and right hepatic arteries originating solely from the proper hepatic artery.25 Successful embolization depends on treatment of all tumor-feeding vessels. In addition, extrahepatic arteries commonly arise from the hepatic arteries, particularly the left. In one study of 250 patients, extrahepatic arterial branches were found in 205 patients, most commonly the right gastric (n = 196), hepatic falciform (n = 129), and accessory left gastric (n = 43) arteries.26 Failure to identify extrahepatic arteries originating from the hepatic arteries can lead to nontarget embolization and morbidity.
Once target arteries are identified, a wire and catheter are advanced into the desired location where embolization is to be carried out. When bilobar disease is present, each lobe is typically treated with independent procedures to reduce the risk of liver failure.5 Patients with tumor burden greater than 75% of the liver volume are at greatest risk for liver failure.27 In these cases, embolization should be performed over several treatment sessions targeting only a small section of the liver at a time.
After embolization is concluded, the catheter is withdrawn and the arterial sheath is pulled. Hemostasis can be obtained with approximately 10 minutes of manual compression or with one of several commercially approved closure devices. Both TAE and TACE result in the postembolization syndrome of elevated liver enzymes, abdominal pain, nausea/vomiting, and fever. This typically resolves in 48 to 72 hours. For this reason, most patients are admitted to the hospital for at least one night following embolization for symptom management. Transarterial radioembolization typically causes milder symptoms and is more commonly performed on an outpatient basis.
Transarterial embolization, or “bland” embolization, involves the transarterial administration of embolic material without chemotherapeutic or radiotherapeutic agents (Fig. 130-2). The goal of this therapy is to cause ischemic tumor necrosis via terminal arteriolar occlusion. TAE for the treatment of carcinoid syndrome was first described in the early 1980s utilizing gelatin foam or powder as the embolic agent.28-30 Since that time, embolic agents have evolved to include polyvinyl alcohol (PVA) particles and size-calibrated microspheres.31,32 These tightly calibrated microspheres have the theoretic advantage of providing better terminal arteriolar blockade.
A.Contrast-enhanced CT of a 71-year-old man demonstrating extensive pancreatic neuroendocrine liver metastases including a dominant segment 8 lesion (arrow). B. Selective celiac angiogram performed during the bland embolization procedure. C. Selective right hepatic artery angiogram during bland embolization procedure. D. Postembolization angiogram demonstrating embolic material and near stasis (arrow) within the targeted anterior branch of the right hepatic artery. E. Contrast-enhanced CT obtained 3 months post-bland embolization demonstrating a necrotic segment 8 tumor (arrow).
Strosberg et al33 retrospectively reviewed the outcomes of 84 patients with NET liver metastasis treated with a total of 161 bland embolizations between 2002 and 2004. Embolization was performed with either 250- to 355-μm PVA particles or 500- to 700-μm microspheres. Of 55 patients who reported symptoms of carcinoid syndrome prior to embolization, 44 (80%) experienced an improvement of symptoms. There was no procedurerelated mortality. Survival data from this and other series are included in Table 130-1.
Recent Survival Data for Various Transarterial Therapies
|Favorite Table|Download (.pdf) TABLE 130-1:
Recent Survival Data for Various Transarterial Therapies
|Study; Year ||Patients ||Therapy ||Survival Data ||Comments |
|Gupta et al27 ; 2005 ||123 patients ||TACE (n = 49)/TAE (n = 74) || |
1-, 2-, and 5- survival rates: carcinoid tumors 95.3%, 68.6%, and 28.6%, respectively; pNET 68.8%, 48.7%, and 13.7%, respectively
|Comparing TACE to TAE, addition of chemotherapy did not improve outcomes in carcinoid patients, but had trend towards improved outcomes for pNET. |
|Strosberg et al33; 2006 ||84 patients ||TAE || ||Median overall survival significantly longer for carcinoid vs. pNET (44 vs. 31 months) |
|Dong and Carr55; 2011 ||123 patients ||TACE || |
3-, 5-, and 10-year overall survival rates: 59%, 36%, and 20%, respectively
Mean survival 3.3 years
|Average 7.3 TACE treatments |
|Cao et al56; (2010) ||58 patients || 90Y Radioembolization || |
1-, 3-, and 5-year survival rate: 86%, 58%, and 47%, respectively.
Median survival 36 (1–61) months
|Extent of tumor, radiographic response, extrahepatic disease, and tumor grade were significant prognostic factors |
|Paprottka et al57 ; 2012 ||42 Patients || 90Y Radioembolization || ||36 of 38 patients with symptom improvement |
|Memon et al43; 2012 ||40 patients || 90Y Radioembolization || |
1-, 2-, and 3-year overall survival rates: 72.5%, 62.5%, and 45%, respectively.
|EASL (complete response, 20.5%; partial response, 43.4%) |
In 2012, Lewis et al34 reviewed complications following 174 TAEs for carcinoid (n = 112) and islet cell (n = 62) liver metastases. The median hospital stay was 4 days (range 1 to 8 days). Thirty-seven percent of patients experienced postembolization fever and there was one case of hepatic abscess occurring 1 month after embolization. Two cases of carcinoid crisis occurred within 48 hours of treatment. One of these patients had been pretreated with octreotide, the other had not. Nine patients required ICU transfers for myocardial infarction (n = 1), Takatsubo cardiomyopathy (n = 1), and hypertensive crisis (n = 7). There was one in-hospital death associated with portal venous gas. The 30-day mortality rate was 0.5%.
Conventional Transarterial Chemoembolization
Conventional transarterial chemoembolization (cTACE) involves the transarterial administration of chemotherapy in combination with embolization. The chemotherapy is suspended in an iodized oil to improve its concentration in the tumor bed. Embolization has the dual effect of limiting chemotherapy washout and causing hypoxic injury. This technique allows for a much higher local concentration of chemotherapy than would otherwise be tolerated with systemic administration and has a longer dwell time than with transarterial chemoinfusion (without embolization). Various chemotherapeutic agents have been used alone and in combination including cisplatin, vinblastine, streptozocin, 5-FU, doxorubicin, and mitomycin C.5 There is no controlled comparison data to support the use of one chemotherapeutic agent to another. A common combination used in the United States is cisplatin, doxorubicin, and mitomycin C.
In 2013, Arrese et al35 examined the role of cTACE in patients with extrahepatic metastases. This retrospective study involved 192 patients with neuroendocrine liver metastases treated with cTACE between 1992 and 2008. Of these patients, 123 patients had evidence of extrahepatic metastases at the time of cTACE. Despite having similar biochemical and radiologic responses as the group without extrahepatic metastases, the group with extrahepatic metastases showed an even greater symptomatic response to treatment (79% vs. 60%, p = 0.01). The authors concluded that cTACE has a role in symptomatic management of NET liver metastases even in the setting of extrahepatic disease.
Drug-Eluting Bead Transarterial Chemoembolization
Drug-eluting beads (DEBs) are size-calibrated micropheres which can be loaded with chemotherapeutic agents for use in chemoembolization (DEB-TACE). When compared to cTACE, DEB–TACE has a favorable pharmacologic profile with sustained delivery of chemotherapy to the tumor bed and decreased systemic peak plasma levels.36 In the management of hepatocellular carcinoma (HCC), several studies have demonstrated an improved tumor response and better safety profile when compared to cTACE.37,38
However, two recent studies have found an increased risk of bile duct injury with DEB-TACE for NET liver metastases not seen in studies evaluating DEB-TACE for HCC. In one phase II study, 7 of 13 patients (54%) treated with doxorubicin-eluting beads for NET liver metastases developed bilomas, a rate much higher than expected. This led to interruption of the trial and modification of technique.39 Similarly, Guiu et al40 evaluated 208 patients with either NET liver metastases (n = 120) or HCC (n = 88) treated with either cTACE or DEB-TACE. Biloma and/or parenchymal infarct was significantly associated with both NET (OR: 8.1; p = 0.04) and DEB-TACE (OR = 9,78; p = 0.002). The authors cautioned the use of DEB-TACE for NET metastases.
Transarterial radioembolization involves the administration of glass or resin embolic particles labeled with yttrium-90 (90Y) (Fig. 130-3).90Y emits beta radiation which travels approximately 10 mm within the liver. With transarterial administration, a dose upwards of 150 Gy can be safely administered to the target lesions. Comparatively, external beam radiation at doses greater than 40 Gy carries a high risk of severe radiation-induced liver disease.41
A. Arterial phase contrast-enhanced CT of a 53-year-old man with pancreatic neuroendocrine tumor demonstrating extensive bilobar liver metastases including a reference lesion (arrow) which measured 3.5 cm prior to ablation. B. Celiac angiogram performed during the mapping procedure in preparation for90Y radioembolization. Note the bilobar tumor blush as well as coil embolization of right gastric artery (arrow) performed to prevent nontarget radiation injury. C. Selective right hepatic angiogram performed during right90Y radioembolization. D. Selective left hepatic angiogram performed during left90Y radioembolization. This procedure was performed 1 month following right radioembolization. E. Arterial phase contrast-enhanced CT performed 6 months post90Y radioembolization demonstrating decreased bilobar metastases. The reference lesion (arrow) decreased from 3.5 to 2.6 cm.
Prior to90Y radioembolization, a planning angiography procedure is performed for dosimetry calculations and prevention of nontarget radiation injury to the lungs and abdominal viscera. During this initial procedure, hepatic arterial variants are identified and extrahepatic visceral branches such as the left gastric and gastroduodenal arteries are occluded, if necessary, to prevent reflux of particles during treatment. Next, radiolabeled macroaggregated albumin (99m Tc-MAA) is injected transarterially for the calculation of the lung shunt fraction.
These measurements are used in dosimetry calculations in order to divide the dose among the various feeding arteries and to limit lung exposure to less than 30 Gy per single session or 50 Gy cumulative.42
In 2012, Memon et al43 reported on the long-term outcomes of 40 patients with NET liver metastases treated with90Y between 2003 and 2007. Of the 25 patients with symptoms at baseline, 21 (84%) patients reported improvement of symptoms. There were no treatment-related deaths. There was one incident of radiation-induced cholecystitis requiring subsequent cholecystectomy. The 1-, 2-, and 3- year survival rates were 72.5%, 62.5%, and 45% after treatment.
In 2009, Riaz et al44 performed a comprehensive literature review of complications following 90Y radioembolization. A postradioembolization syndrome including fatigue, abdominal discomfort, and/or nausea and vomiting occurred in a range of 20% to 55% of cases, sometimes requiring symptomatic management. Radiation-induced liver disease was reported at a rate of 0% to 4%, most commonly in patients with preexisting liver dysfunction. GI ulceration from nontarget radioembolization was reported to be less than 5%. The incidence of radiation-induced cholecystitis and radiation pneumonitis was each less than 1%. The authors concluded that while postradioembolization syndrome is relatively common, hospitalization is rarely necessary for symptom management. The more serious adverse events can typically be prevented using pretreatment dosimetry and meticulous pretreatment mapping.
Comparison of Transarterial Therapies
While there is a lack of prospective randomized trials comparing the various transarterial therapies, several studies have compared modalities retrospectively. Gupta et al27 retrospectively reviewed the outcomes of 123 patients with neuroendocrine liver metastases treated with either TAE (n = 74) or TACE (n = 49). For the subset of patients with carcinoid histology (n = 69), TAE resulted in a higher CR + PR response rate compared to TACE (TAE 81.0% vs. TACE 44.4%, p = 0.003). For the subset with pancreatic neuroendocrine liver metastases (n = 54), TACE demonstrated a trend toward improved response that was not statistically significant (TAE 25.0% vs. TACE 50.0%, p = 0.06).
The authors also reviewed complications for both TAE and TACE. Serious adverse events were documented in 25 embolizations (8.5%). There were seven cases of hepatorenal syndrome and six cases of sepsis. Two patients required drainage of hepatic abscesses. Other complications included transient myelosuppression (n = 1), anasarca (n = 1), transient cortical blindness (n = 1), and necrotizing cholecystitis that required cholecystectomy (n = 1). There was one death secondary to myocardial infarction, which occurred 3 days after embolization. The complication rates for the TACE and TAE groups were 20% and 12%, respectively (p = 0.31).
In 2008, Pitt et al45 published a retrospective analysis of 100 patients treated with either TACE (n = 49) or TAE (n = 51) at three institutions between 1996 and 2007. Of the 100 patients, 56 patients had carcinoid histology and 44 had pancreatic NETs. No subset analysis was performed for these histologies. The groups treated with TACE and TAE had similar median survival calculated from the time of metastasis diagnosis (50.1 vs. 39.1 months, respectively; p = 0.62) and calculated from the time of first embolization (25.5 vs. 25.7 months, respectively; p = 0.79). Of the patients with tumor-related symptoms, the percentage who responded to treatment were similar between TACE and TAE (86% vs. 83%, respectively; p = NS). The authors concluded that TACE and TAE offer similar rates of morbidity, symptom improvement, and overall survival for the management of neuroendocrine liver metastases.
In 2012, Yang et al46 performed a systematic review comparing 90Y radioembolization to TAE/cTACE/DEB-TACE in managing neuroendocrine liver metastases. Of the 37 studies included, 12 were 90Y (423 patients) and 25 were other transarterial therapies (11 TACE, 8 TAE, 3 TACE or TAE, and 1 DEB-TACE, total 1152 patients). Pooled data comparing 90Y and TAE/TACE/DEB-TACE showed similar objective response (63.1 vs. 58.4%, respectively), clinical response (85.0% vs. 88.5%, respectively), and median survival (28 months vs. 34.9 months, respectively). The authors concluded that the available retrospective data show similar efficacy between 90Y and the other transarterial therapies and that prospective multi-institutional comparative studies are warranted.
Given the present state of data, the decision to employ one transarterial modality over another is made on a case by case basis taking into account various factors. For instance, because 90Y particles are less embolic than the other transarterial therapies, there is a theoretic benefit to this therapy in the setting of advanced liver failure or portal vein thrombus. Other factors include local practice standards and operator preference/experience. Cost is another consideration as at least two cost analysis studies have demonstrated 90Y to be a generally more expensive option than other transarterial therapies.47,48