Nonsurgical local treatment strategies for colorectal cancer liver metastases

INTRODUCTION — Surgical resection is the treatment of choice for patients with isolated colorectal cancer (CRC) liver metastases when feasible. For patients with four or fewer isolated hepatic lesions, five-year relapse-free survival rates after resection range from 24 to 58 percent, averaging 30 percent (table 1). Many of these patients are potentially cured. However, even with liver-limited disease, the majority of patients are not surgical candidates because of tumor location, multifocality, or inadequate hepatic reserve.

There are several nonsurgical treatment options for patients with liver-isolated CRC metastases who are not candidates for potentially curative resection. These include systemic chemotherapy, as well as liver-directed therapies such as:

●Tumor ablation (radiofrequency or microwave coagulation, interstitial laser ablation)

●Radiation therapy, which includes conventional fractionation external beam radiation therapy and stereotactic body radiation therapy

●Regional chemotherapy via the hepatic artery

●Radioembolization using yttrium-labeled glass or resin microspheres

Some of these techniques may also be useful for patients with metastatic CRC and isolated or predominant liver metastases who have become refractory to systemic chemotherapy.

This topic review will focus on indications and techniques for nonsurgical liver-directed methods for local tumor ablation, regional chemotherapy into the hepatic artery, radioembolization and radiotherapy. Resection and the use of pre-resection (induction) chemotherapy, as well as systemic treatment strategies for metastatic CRC, are addressed elsewhere. (See "Potentially resectable colorectal cancer liver metastases: Integration of surgery and chemotherapy" and "Systemic therapy for metastatic colorectal cancer: General principles" and "Systemic therapy for nonoperable metastatic colorectal cancer: Selecting the initial therapeutic approach".)

CHEMOTHERAPY-NAIVE DISEASE

Surgical candidates — Surgical resection is the treatment of choice for patients with isolated, potentially resectable CRC liver metastases when feasible. Surgical therapy in patients with liver-isolated disease offers a substantial survival benefit compared with no surgery. At present, it is uncertain that nonsurgical treatment options offer equivalent benefit to surgical resection.

Although surgical resection is typically the treatment of choice for patients with isolated, potentially resectable CRC liver metastases, some patients who are surgical candidates may be offered initial systemic chemotherapy with deferred resection:

●In the setting of synchronous metastatic disease, a major advantage of this approach is that it allows an understanding of the natural history of the metastatic disease before subjecting the patient to surgery that might not be curative. (See "Systemic therapy for metastatic colorectal cancer: General principles", section on 'Potentially resectable disease'.)

●Patients with isolated CRC initially unresectable but potentially resectable liver metastases, who might be considered surgical candidates if their metastases were smaller. (See "Potentially resectable colorectal cancer liver metastases: Integration of surgery and chemotherapy", section on 'Patients with initially unresectable metastases' and "Hepatic resection for colorectal cancer liver metastasis", section on 'Patient selection'.)

An unanswered question is whether any form of local ablation is as good or better than resection for patients with potentially resectable CRC liver metastases is not established; there are no adequately powered high-quality randomized trials that have been conducted, although at least one such trial, the COLLISION trial (NCT03088150) [1] comparing radiofrequency ablation (RFA) versus resection, is currently accruing patients.

While many local nonsurgical methods (particularly RFA, microwave ablation [MWA], and stereotactic body radiotherapy) can provide prolonged local control in many patients, there are fewer long-term survivors (ie, cures) compared with surgical resection, perhaps as a function of the clinical circumstances that led to a nonsurgical intervention in the first place. As a result, surgery is generally preferred for a patient with potentially operable CRC liver metastases, if a margin-negative resection can be accomplished.

Information is available from three systematic reviews [2-4]; the most recent compared RFA or MWA versus partial hepatectomy in the treatment of CRC liver metastases [2]; it included predominantly observational comparative studies (some utilizing retrospective data collection), and the one randomized trial of MWA versus resection (described below [5]) was downgraded for quality because of serious concerns about risk of bias [2]. (See 'Radiofrequency ablation' below.)

The authors concluded that:

●Resection appeared superior to RFA alone (10 observational comparative studies, hazard ratio [HR] for death 1.78, 95% CI 1.35-2.33), but RFA was associated with significantly fewer complications (10 observational comparative studies, HR for complications 0.47, 95% CI 0.28-0.78). Survival rates were similar for resection and MWA based upon one randomized trial with serious concerns about risk of bias.

●RFA plus resection was not significantly better than resection alone (7 observational comparative studies, HR for death 1.24, 95% CI 0.84-1.84).

However, these data are less than definitive given the retrospective nature of many of the included studies. One problem with nonrandomized series such as those included in this analysis is that the lower rates of long-term survivors (ie, cures) with ablation compared with surgical resection may be unrelated to the procedure, and instead, a function of the clinical circumstances that led to a nonsurgical intervention in the first place. Furthermore, the risk of complications from ablation procedures is dependent on the size and location of the lesion, which was not provided.

Unsuitable for resection

Liver-directed versus systemic therapy alone — For patients with isolated liver metastases and a limited number of small lesions who are felt to be unsuitable for resection because of tumor location, impaired general health status, or an insufficient future liver remnant to resect all lesions, locoregional liver-directed treatments such as ablation or SBRT are an alternative to initial systemic chemotherapy. One advantage of locoregional liver-directed therapy, for appropriately-selected patients, is deferral of chemotherapy-related toxicity. However, systemic chemotherapy is generally well tolerated if close attention is paid to avoid dose-limiting oxaliplatin neuropathy, and is a reasonable alternative option, particularly if local expertise to perform nonsurgical regional therapy is not available, and patients have not received prior adjuvant chemotherapy. Patients who first receive systemic chemotherapy may be candidates for "consolidation treatment" of residual masses with ablation if disease remains liver limited. (See 'Local ablation plus systemic chemotherapy' below.)

The relative benefit of locoregional liver-directed therapies over modern palliative combination systemic chemotherapy alone for patients who are unsuitable for resection is uncertain. There are no randomized trials in which patients with chemotherapy-naïve liver isolated hepatic metastases who were unsuitable for resection were randomly assigned to any form of liver-directed therapy versus initial systemic therapy. Furthermore, most of the studies examining the different forms of locoregional liver-directed therapy were conducted at a time before the introduction of oxaliplatin, irinotecan, and biologic agents such as bevacizumab and agents targeting the epidermal growth factor receptor. Long-term survival for patients with metastatic CRC has clearly improved with the availability of more active anticancer agents. As an example, in a report of pooled data from North Center Cancer Treatment Group trials conducted in the fluorouracil (FU) plus leucovorin (LV) era, only 1.1 percent of patients were alive at five years [6]. By contrast, in a report from the phase III FIRE-3 trial (first-line irinotecan with short-term infusional FU plus LV [FOLFIRI] plus either bevacizumab or cetuximab), the five-year survival rate for patients with RAS wild-type tumors treated with FOLFIRI plus cetuximab was approximately 20 percent [7]. An important point is that although many of the survival gains are attributable to advances in chemotherapy treatment, more aggressive use of surgical resection of metastatic disease has also contributed. (See "Systemic therapy for metastatic colorectal cancer: General principles", section on 'Systemic therapy versus supportive care'.)

In general, these benefits are obtained with sequential use of all active agents, and it is still not established whether there is any survival benefit to initiating systemic chemotherapy early, at the time metastatic disease is diagnosed, versus deferring therapy until patients become symptomatic. This subject is discussed elsewhere. (See "Systemic therapy for metastatic colorectal cancer: General principles", section on 'Timing of systemic therapy'.)

A different question is whether the addition of locoregional liver-directed therapies to systemic chemotherapy improve outcomes, this is addressed below. (See 'Local ablation plus systemic chemotherapy' below.)

do not improve survival over systemic chemotherapy alone, and for most patients who are responding to palliative systemic chemotherapy for metastatic CRC, we suggest not pursuing combined therapy. (See 'Tumor ablation plus systemic chemotherapy' below and 'Radioembolization plus systemic chemotherapy' below.)

Options for liver-directed therapy — The options for locoregional liver-directed therapy are tumor ablation or stereotactic body radiotherapy (SBRT). If the decision is made to perform tumor ablation, for most patients, we suggest RFA or MWA rather than hepatic intra-arterial chemotherapy or radioembolization. The best results with RFA are in patients with three or fewer lesions that are each less than 3 cm in diameter and are not located near major vascular structures. Where available, MWA is a reasonable alternative to RFA, especially for tumors located near vascular structures. (See 'Tumor ablation' below.)

The benefit of SBRT relative to tumor ablation is unclear, and the choice usually depends on local expertise and patient preference. (See 'Stereotactic body radiotherapy' below.)

Hepatic intra-arterial chemotherapy delivered via implanted pump is a therapeutic option for patients with isolated liver metastases that are not amenable to surgical resection or other forms of local ablation, but it is not widely used and its place in the therapeutic armamentarium for liver-isolated metastatic CRC is not established. If it is used, this approach should be restricted to centers with expertise in the technical aspects of regional chemotherapy. (See 'Hepatic intra-arterial chemotherapy' below.)

Tumor ablation — For patients with isolated liver metastases from CRC who are not suitable for potentially curative resection because of tumor location, impaired general health status, or an insufficient future liver remnant to resect all lesions, locoregional liver-directed methods of tumor ablation provide an alternative to palliative systemic chemotherapy. (See "Systemic therapy for nonoperable metastatic colorectal cancer: Selecting the initial therapeutic approach".)

Choice of technique — Hyperthermic ablation (using radiofrequency waves or microwaves) has largely become the ablation method of choice based on its safety profile and results, particularly in patients with three or fewer metastases with none greater than 5 cm in diameter.

RFA and MWA can be administered as an outpatient procedure and may be considered a less morbid alternative to surgical resection in patients who are at high risk for, or otherwise not candidates for, surgery [8]. Hyperthermic ablation using RFA or MWA is limited by the difficulty in determining the true lesion margin, which often extends beyond the leading edge, and the lack of specificity of tissue damage. The risk of local recurrence after ablation is increased in lesions larger than 3.5 to 4.0 cm.

Whether MWA is more effective or safer than RFA is not certain; there are no published randomized trials. Some retrospective comparative series have concluded that RFA might be preferable for peribiliary lesions, while others suggest that MWA may be a better choice for perivascular tumors [9,10].

A systematic review of 75 studies on thermal ablation for CRC liver metastases, most studies (36) reported results for RFA, with local recurrence rates ranging from 10 to 31 percent, whereas studies reporting on outcomes after MWA (n = 13) had a much lower local recurrence rate (5 to 13 percent) [11].

A later meta-analysis of observational studies comparing MWA versus RFA for a variety of hepatic lesions (16 studies involving 2062 patients) concluded that while MWA was associated with a better 6-year overall survival, this was based on only a few series [12]. Other outcomes (one- and five-year overall survival, local recurrence rate, adverse effects) were similar.

There are numerous methods to achieve tumor ablation, and this clinical problem has drawn much interest from the biomedical technology world. Cryotherapy (tumor freezing) was deemed too cumbersome and was succeeded by other modalities including the direct intratumoral instillation of alcohol or acetic acid.

Radiofrequency ablation — RFA has been widely applied to patients with primary hepatocellular cancer (HCC) and metastatic liver tumors, though differences in surrounding liver (eg, cirrhosis) and tumor physical characteristics should preclude extrapolation of results from one tumor type to the other. RFA can be performed using open, laparoscopic, or percutaneous approaches. Some studies have reported that the approach by which RFA is performed has an impact on tumor recurrence rates, with the fewest local recurrences after open RFA, followed by laparoscopy, and finally percutaneous RFA [13-16]. However, local tumor recurrence rates overlap broadly with each technique, and clinician experience (as well as the type of RFA equipment) is also inversely related to local recurrence rates [17,18]. An expert panel convened by the American Society of Clinical Oncology (ASCO) to review the evidence on RFA for CRC liver metastases concluded that there is insufficient evidence to resolve the issue of optimal approach [3].

Several factors limit the success rate of RFA, including lesion size, location, and lesion number:

●Lesion size and RFA success – The highest ablation success rates are achieved in patients with a solitary metastasis or a few metastases that are all less than 3 cm in size [19-22]. The relationship between size and local recurrence rate is likely related to the size of the thermal injury that can be created by modern RFA devices [23]. Because the objective of thermal ablation is to induce coagulative necrosis of the targeted tumor as well as a rim of normal hepatic parenchyma (5 to 10 mm), the best results are obtained when the tumor is smaller than the size of the coagulative necrosis produced by a single ablation probe (currently about 4 cm). In most (but not all [24]) series, local failure rates increase dramatically when tumor size increases, especially beyond 3.5 to 4 cm [11,25-29]. As an example, in one study, local tumor progression rates after RFA for lesions ≤3 cm versus >3 to 5 cm were 15 and 61 percent, respectively [29].

●Impact of tumor location and number – The location of the tumors within the liver also impacts on the success of RFA:

•Lesions that are located near large (≥1 cm) blood vessels may be inadequately treated because of the heat sink effect of rapid blood flow [25,30,31].

•Percutaneous RFA may sometimes be avoided for treatment of lesions that are located in the dome or along the inferior liver edge for fear of diaphragmatic injury [32] or intestinal perforation. However, with attention to technique (such as insulating adjacent bowel from the thermal process, an approach that can be accomplished using a laparoscopic or open laparotomy approach), these lesions can be successfully treated [33]. Tumors at the inferior edge of the liver should not be treated percutaneously (and instead by a laparoscopic or open approach) if the stomach, duodenum, or transverse colon lies in close proximity to the tumor. Hydrodissection is a technique whereby installation of saline between the target tumor and adjacent structures (eg, diaphragm) is used to protect the adjacent structure from thermal injury during tumor ablation.

•Ablation of solitary metastases is associated with very high rates of local tumor control and survival (5-year survival >50 percent in two separate series [34,35]).

●Long-term outcomes – The vast majority of published data on efficacy of RFA for CRC liver metastases come from retrospective series, many of which have limited follow-up (20 months or less); there are few published randomized trials [13,36-41]. A systematic review of the literature on RFA for CRC liver metastases reported a wide range of five-year survival (14 to 55 percent) and local recurrence rates (3.6 to 60 percent) [3]. However, both the retrospective series and limited number of prospective trials consist of a variable mix of patients with potentially resectable liver-isolated disease and unresectable liver metastases with or without extrahepatic disease involvement.

Among appropriately selected patients with potentially resectable liver metastases [42-48], at least three systematic reviews and meta-analyses have been conducted [2-4], with all three concluding that particularly in the absence of extrahepatic disease, there is not enough evidence to support the use of RFA over resection in patients with potentially resectable CRC liver metastases. (See 'Surgical candidates' above.)

●Complications – RFA is a relatively well-tolerated technique; however, severe and potentially fatal complications can arise. In the ASCO systematic review, the reported mortality rate was 0 to 2 percent, and the major complication rate was between 6 and 9 percent in most studies [3]. Postablation syndrome, which is characterized by low-grade fever, malaise, chills, delayed pain, and nausea and vomiting, occurs in 30 to 40 percent of patients, and appears, on average, three days after ablation and lasts five days [49]. It is usually self-limiting and resolves within ten days. Opioids are rarely needed.

Other postprocedure complications may be more serious, and not self-limiting. One of the largest series to address other potentially more serious complications included 312 patients with hepatic tumors (predominantly colorectal metastases) who underwent 350 procedures (226 percutaneous and the remainder intraoperative) [50]. Five deaths were attributed to treatment (one each from liver failure and colon perforation, three from portal vein thrombosis). Portal vein thrombosis was significantly more common in cirrhotic (2 of 5) compared with noncirrhotic livers (0 of 54) after intraoperative RFA performed during a Pringle maneuver. There were 37 nonfatal serious complications (incidence 10.6 percent), which included:

•Liver abscess in seven, which developed in all three patients with a bilioenteric anastomosis compared with less than 2 percent of the others

•Pleural effusion and skin burns in five patients each

•Hypoxemia during treatment in four patients

•Pneumothorax in three patients

•Subcapsular hematoma in two patients

•Acute renal insufficiency, hemoperitoneum, and needle tract seeding in one patient each

Microwave ablation — Microwave systems use an electromagnetic signal to generate heat by vigorous agitation of water molecules in tissues, resulting in cellular death by coagulation necrosis [5,51].

Although best studied for treatment of small HCCs, MWA has gained acceptance as a favorable alternative to RFA, and in some cases, preferred over RFA, for thermal ablation of metastatic liver tumors [9,10,52-54]. There are several advantages of MWA over RFA, such as higher intratumoral temperatures, incorporation of multiple applicators simultaneously, allowing treatment of multiple lesions, and less procedural pain [55,56]. Because the electromagnetic waves produce thermal necrosis by molecular friction, the heat sink effect is less of an issue with MWA than with RFA. Although MWA can be performed intraoperatively [57], at most institutions, MWA is now performed percutaneously by interventional radiologists [53,58].

●MWA versus resection – In one of the only randomized trials directly comparing local nonsurgical ablation versus resection, 30 patients with potentially resectable hepatic metastases from CRC were randomly assigned to laparotomy with ultrasonographically-guided microwave ablation (MWA) or surgical resection [5]. Both approaches were associated with similar two-year (56 and 57 percent) and three-year (14 and 23 percent) survival rates, as well as similar median survival (27 and 25 months, respectively). Notably, this trial was downgraded for quality because of serious concerns about risk of bias in one meta-analysis [2]. (See 'Surgical candidates' above.)

Additional information is available from a propensity score analysis derived from a population-based Swedish registry of all patients undergoing liver resection or MWA as a first intervention for CRC liver metastases measuring ≤3 cm between 2013 and 2016 [59]. The characteristics of the unmatched cohort of 82 MWA patients and 645 resected patients differed significantly as to age, comorbidity, tumor location, number of metastases, and prior chemotherapy, and overall survival favored resection (three-year overall survival 76 versus 69 percent). However, a difference was no longer evident in cohorts with propensity score matching (70 MWA and 201 resected, three-year overall survival 76 versus 76 percent).

Long-term outcomes with MWA for CRC liver metastases are further supported by several retrospective series [60,61].

●Complications – Reported complications of percutaneous MWA are similar to those reported for RFA, and are typically mild, including pain, fever and transaminase elevation; the risk of liver abscess, bile leak/biloma, ascites/pleural effusion, diaphragm injury, and needle track seeding are all low [53,60].

Interstitial laser ablation — Published experience with interstitial laser thermotherapy is limited to a few institutions [58,62-64]. The largest series consisted of 603 patients who underwent magnetic resonance imaging (MRI)-guided laser thermotherapy of 1801 CRC liver metastases; details of additional oncologic therapy were not provided [64]. Clinically relevant complications included pleural effusion (1.1 percent), liver abscess (0.4 percent), and intra-abdominal bleeding, pneumothorax, bile duct injury, bronchobiliary fistula, and death within 30 days in 0.1 percent each. At six months postprocedure, 95 percent of the lesions were controlled locally, and none recurred thereafter with up to 7.6 years of follow-up; median follow-up duration was not stated. Five-year survival was 37 percent.

Response assessment — The efficacy of percutaneous ablation therapy is typically assessed by periodic contrast-enhanced dynamic computed tomography (CT) or MRI starting one month after therapy. Standard methods to assess treatment response to anticancer therapies (eg, unidimensional Response Evaluation Criteria in Solid Tumours [RECIST] criteria (table 2) [65], bidimensional perpendicular measurement using World Health Organization [WHO] criteria [66]) involve measurement of tumor dimensions before and after treatment, and categorize the degree of response according to the degree of shrinkage. However, methods such as these disregard the extent of necrosis, which is the end result of locoregional ablative therapies, and may underestimate the disease response posttreatment.

In patients with HCC, the degree of contrast enhancement within tumor masses is an indicator of viable tumor, and methods to assess response to local ablation are based on the persistence or growing areas of contrast enhancement on cross sectional imaging rather than changes in tumor size. (See "Assessment of tumor response in patients receiving systemic and nonsurgical locoregional treatment of hepatocellular cancer".)

However, for patients with CRC liver metastases, contrast enhancement is typically absent, and there is no consensus on appropriate endpoints to define clinical benefit from locoregional therapies. Treated tumors characteristically show low density on posttreatment CT scans. These regions, which are often interpreted as tumor necrosis, may even be larger than the area of the original tumor. However, this appearance is nonspecific and should not be considered to represent an "objective response," and these response categories should not be used. The most important indicator of benefit from these therapies is the time to disease progression as detected by periodic cross sectional imaging. Disease progression may be manifest by growth of tumor at the treated site, progression in other parts of the liver, or in extrahepatic sites.

Some patients have elevated levels of serum tumor markers such as carcinoembryonic antigen (CEA). While a reduction in serum CEA levels may not be accompanied by objective tumor shrinkage in patients treated with percutaneous ablation therapy, at least some data suggest a correlation between the magnitude of CEA decline and survival following ablation [67]. As with patients treated with systemic chemotherapy, a rising CEA in a patient treated with local ablation may precede radiographic tumor progression. Treatment decisions should not be based on rising tumor markers levels alone, but progressive rising tumor markers should prompt repeat radiographic reevaluation. (See "Systemic therapy for metastatic colorectal cancer: General principles", section on 'Assessing treatment response'.)

Stereotactic body radiotherapy — There are no randomized trials comparing SBRT with other forms of liver-directed nonsurgical therapy, and its role in the therapeutic armamentarium for unresectable liver metastases, particularly in patients who are eligible for thermal ablation (RFA, MWA), is not well defined. Nevertheless, SBRT is an effective and safe alternative to thermal ablation, with durable local control. In practice, the choice of SBRT over hyperthermic ablation using RFA or MWA generally depends on local expertise and patient preference. One issue is that RFA or MWA can often be carried out in one treatment session, while SBRT often requires several sessions. SBRT may also be preferred over thermal ablation for lesions that abut a large blood vessel.

SBRT is a technique that utilizes precisely targeted radiation to a tumor while minimizing radiation to adjacent normal tissue. This targeting allows treatment of small- or moderate-sized tumors in either a single or limited number of dose fractions. Early experience with SBRT for liver metastases suggests that this technique is safe and associated with sustained local control, although the reported range of long-term local control is broad (59 to 91 percent at two to three years) [68-74]. Post SBRT local control is generally defined as disappearance of, decrease in, or no increase in the size of the treated lesion(s). (See "Radiation therapy techniques in cancer treatment", section on 'Stereotactic radiation therapy techniques'.)

As examples:

●As noted above, in a systematic review of 18 retrospective series encompassing 656 patients treated with SBRT for CRC liver oligometastases (most studies specified no more than two liver metastases, and most patients had received at least one or two lines of systemic chemotherapy), the pooled local control rates at one and two years (available from 13 studies), were 67 and 59 percent, respectively [72]. Pooled rates of grade 1 to 2 and 3 to 4 acute liver toxicity were 30.7 and 8.7 percent, respectively. (See 'Liver-directed versus systemic therapy alone' above.)

●Comparable results were noted in a phase II trial of SBRT for unresectable liver metastases in 61 patients, of whom 29 had a CRC primary [73]. In this group, the local control rate at three years was 75 percent, and results were not significantly different for lesions smaller or larger than 3 cm (77 versus 81.9 percent, respectively).

●In a registry-based analysis of 217 patients with 233 CRC liver metastases treated with SBRT and reported to the international multi-institutional RSSearch Patient Registry between 2005 and 2017 (median follow-up 15 months), one- and two-year local control rates were 75 and 55 percent, respectively [74].

Largely based on the systematic review, year 2022 guidelines for managing CRC liver metastases from the ASCO recommended consideration of SBRT following systemic therapy for patients with oligometastases of the liver who are not considered candidates for resection [75]. (See 'Tumor ablation plus systemic chemotherapy' below.)

Hepatic intra-arterial chemotherapy — Despite a multitude of trials, the relative benefit of regional (hepatic arterial infusional chemotherapy [HAIC]) over systemically-administered chemotherapy for patients with hepatic metastases from CRC remains unclear, particularly within the context of newer systemic therapies that include oxaliplatin, irinotecan, and the biologic agents (bevacizumab, anti-epidermal growth factor strategies). Regional chemotherapy through the hepatic artery remains a therapeutic option for patients with isolated liver metastases that are not amenable to surgical resection or local ablation, but it is not widely used, and its place in the therapeutic armamentarium for metastatic CRC is not established. If it is used, this approach should be restricted to centers with expertise in the technical aspects of regional chemotherapy.

HAIC is based upon the following principles:

●Liver macrometastases (>0.5 cm) derive more than 80 percent of their blood supply from the hepatic arterial circulation, while normal hepatocytes are supplied primarily by the portal circulation [76]. As a result, the administration of chemotherapy into the hepatic artery allows the selective delivery of drug to the tumor with relative sparing of normal hepatocytes [77].

●Depending upon a drug's clearance and toxicity profile, a marked increase in the local concentration of drug may be achieved by injection into the hepatic artery [77]. Regional administration of drugs that are rapidly metabolized in the liver by a first-pass effect leads to higher levels of drug exposure and minimizes systemic side effects.

Extensive clinical investigation of HAIC began in the 1970s and 1980s, initially via percutaneously placed catheters and then by a totally implantable pump system. In several prospective trials, enrolling over 400 previously untreated patients with metastatic CRC, randomly assigned patients to systemic fluoropyrimidine chemotherapy or to HAIC with FUDR (floxuridine) delivered via an implanted pump [78-82],the response rate to HAIC was consistently superior but this did not translate into a significant survival improvement in any study or in two separate meta-analyses [83,84]. Subsequent improvements in HAIC technique led to a resurgence of interest in the use of this therapy.

●Technique and complications of pump placement – The goal of intrahepatic artery catheter placement is to enable bilobar hepatic perfusion of chemotherapy and prevent administration to the stomach or duodenum (misperfusion).

A preoperative CT angiogram defines the arterial supply of the liver [85]. If suitable for placement of a hepatic artery catheter, and no extrahepatic tumor, a total devascularization of the distal stomach and proximal duodenum is performed in order to minimize the risk of misperfusion [86]. Any variations in the hepatic lobar arterial supply, including a replaced or accessory left or right hepatic artery, require ligation of that vessel [87].

In nearly all cases, the gastroduodenal artery is chosen for cannulation. Once the catheter is situated, the pump itself is placed in a subcutaneous pocket and it is loaded with heparinized water. Dissection of the lymphatics at the porta hepatis is performed to make room for the pump. Finally [85], 5 mL of fluorescein or methylene blue are injected into the pump side port, and the liver, stomach, and duodenum examined to verify bilobar liver perfusion and to exclude misperfusion to the stomach or duodenum.

Postoperatively, before initiating HAIC, a Tc-99m macro-aggregated albumin scan is necessary to ensure the absence of misperfusion and to assess the adequacy of whole liver perfusion [88,89].

Early postoperative complications consist of arterial injury leading to hepatic artery thrombosis, incomplete perfusion of the entire liver due to missed recognition of an accessory hepatic artery, misperfusion to the stomach or duodenum, or pump pocket hematoma. If the above steps are followed rigorously, these complications should occur in fewer than 5 percent of patients.

Late complications include biliary injury, inflammation or ulceration of the stomach or duodenum, pump pocket infection, or catheter thrombosis. The development of antral or duodenal ulceration should prompt endoscopic examination of the stomach and duodenum with concomitant injection of methylene blue through the pump side-port [90]. Immediate deep blue staining of the ulcerated site warrants an angiographic search for a vessel responsible for the misperfusion. Once identified, the offending vessel can often be embolized using interventional radiologic techniques.

When the infusion pump is no longer being used the infusion device needs to be either operatively removed or regularly flushed to prevent thrombosis.

●Optimizing chemotherapy – The dose of fluorodeoxyuridine (FUDR) employed in early phase III trials of HAIC (0.3 mg/kg per day for 14 of 28 days) induced severe biliary toxicity [91]. A number of strategies were explored to maximize the safety and efficacy of HAIC.

•Dexamethasone and leucovorin – In an attempt to ameliorate any inflammatory aspect of the biliary toxicity, FUDR (0.3 mg/kg per day for 14 days) was administered with and without dexamethasone in a randomized phase II trial [92]. While the addition of dexamethasone did not lessen biliary toxicity, both the response rate and median survival were superior to FUDR alone.

A later series of studies tested FUDR modulated by intra-arterial LV (which enhances the activity of FU by promoting the formation of a stable ternary complex with thymidylate synthetase) plus dexamethasone [93]. The response rate in previously untreated patients was 78 percent, and the median survival was 25 months [94]. A strict paradigm for dose-reduction (table 1) reduced the incidence of biliary sclerosis to 3 percent, although dose adjustment or temporary treatment delay was necessary in nearly every patient.

•Oxaliplatin and irinotecan – Oxaliplatin and irinotecan are two highly active agents in metastatic CRC. (See "Systemic therapy for nonoperable metastatic colorectal cancer: Selecting the initial therapeutic approach", section on 'The cytotoxic chemotherapy backbone'.)

Both can be safely delivered to the liver via the hepatic artery [95-98]. In one report, 28 previously untreated patients with liver isolated metastatic CRC received intra-arterial oxaliplatin (100 mg/m² over two hours) followed by IV FU and LV administered according to the de Gramont regimen (LV 200 mg/m² over two hours, followed by FU 400 mg/m² bolus, and then a 22-hour infusion of FU 600 mg/m²), with repeated courses every two weeks [95]. The objective response rate was 64 percent, and the median overall survival was 27 months. Grade 3 or 4 neutropenia complicated therapy in 10 patients, and there were two treatment-related deaths, one in the setting of febrile neutropenia. Whether these results are better than could be achieved by systemic administration of oxaliplatin, FU, and LV (eg, the FOLFOX regimen) is unclear.

●HAIC plus systemic therapy – It was postulated that the combined use of intra-arterial plus systemic chemotherapy might improve outcomes by addressing concerns about extrahepatic tumor progression during hepatic arterial infusional chemotherapy (HAIC) while achieving maximal therapeutic effect in the liver. At least three trials have directly compared HAIC versus systemic chemotherapy, but all used inferior chemotherapy by modern standards in the control group (ie, IV LV-modulated FU alone) [99-101]. A meta-analysis of these three trials combined with seven earlier randomized trials in which the control arm was systemic therapy or best supportive care concluded that response rates were higher with HAIC but this did not translate into better survival [102]. None of the trials included irinotecan or oxaliplatin in the control arm.

The only available data, from a retrospective analysis of a cohort of individuals treated with HAIC plus modern systemic chemotherapy over a 10-year period within a single health care system, suggest that the combined approach of HAIC plus modern combination systemic therapy might be associated with an approximate doubling of median overall survival relative to systemic chemotherapy alone (33 versus 15 months); this was not a randomized trial, and the overall survival in the control group was low by modern standards [103]. Furthermore, the patients in the control group received therapy at affiliated community cancer centers, while all those receiving HAIC had been referred to and managed at the flagship oncologic hospital. It remains unknown whether any HAIC approach will improve upon or even match the results from contemporary modern systemic combination chemotherapy.

Local ablation plus systemic chemotherapy

Tumor ablation plus systemic chemotherapy — The benefit of RFA/MWA or SBRT for regional treatment of liver metastases who are undergoing systemic chemotherapy for chemotherapy sensitive metastatic CRC remains unproven, although recent studies suggest that selected patients may benefit, and it is an option for patients with a limited burden of liver-limited disease following initial systemic therapy. This recommendation is consistent with year 2022 ASCO guidelines on management of metastatic CRC [75].

Local tumor ablation is often used for unresectable CRC liver metastases concurrent with systemic chemotherapy [57], but it has not been clear whether any form of nonsurgical local therapy contributes to better survival, and few, if any, studies have utilized modern systemic combination chemotherapy as the control arm.

There have been no phase III randomized trials of thermal ablation plus chemotherapy versus chemotherapy alone. In the only randomized phase II trial addressing the benefit of combined therapy versus systemic chemotherapy alone, the CLOCC trial, 119 patients with initially unresectable CRC liver metastases and no extrahepatic disease were randomly assigned to systemic treatment alone (oxaliplatin plus LV and short-term infusional FU with or without bevacizumab) or with aggressive local treatment, which was defined as RFA with or without resection [40,104]. Eligibility was limited to patients with fewer than 10 lesions that could all be treated with RFA alone or with combined treatment, which included resection of resectable lesions and RFA of nonresectable lesions.

In the latest analysis, patients randomly assigned to RFA in conjunction with chemotherapy had a significantly longer median survival (45.6 versus 40.5 months), and eight-year overall survival (36 versus 9 percent; HR 0.58, 95% CI 0.33-0.88) [104]. Patients receiving RFA in conjunction with systemic therapy had fewer liver recurrences (65 versus 82 percent), but a higher extrahepatic recurrence rate (35 versus 14 percent). At the RFA sites only 11 of 170 lesions recurred (6.5 percent) in 9 patients (15 percent of the total). Although posttreatment salvage chemotherapy details were not provided, the cause of death was progressive disease in significantly fewer patients who underwent local ablative treatment (57 versus 81 percent).

While this randomized study suggests that combining local ablative and/or surgical resection with systemic chemotherapy improves survival in patients with liver-limited metastatic CRC, one must interpret the results cautiously. The study was closed early due to poor accrual and was conducted in the early 2000s when many of the new therapeutic options were not yet available. Furthermore, in the RFA arm, 27 of the 50 patients also underwent hepatic resection, which may have confounded results. Finally, the overall survival exceeded 40 months in both arms, suggesting extremely unusual patient selection leading to a group of patients with remarkable prognoses.

Radioembolization plus systemic chemotherapy — We suggest against the addition of radioembolization to modern oxaliplatin-containing chemotherapy for treatment of patients with chemotherapy-naïve metastatic CRC. Numerous randomized trials suggest that combined therapy-does not improve survival as compared with chemotherapy alone and is associated with more adverse events. Additional study is needed to define whether there is a subpopulation of patients with liver-isolated or liver-predominant metastatic CRC that might benefit from combined therapy. This recommendation is consistent with year 2022 ASCO guidelines on management of metastatic CRC [75].

A means of delivering focal radiation to the liver employs radioactive isotopes (eg, 131-labeled lipiodol or yttrium 90-tagged glass or resin microspheres) that are delivered selectively to the tumor via the hepatic artery (radioembolization, also called selective internal radiotherapy [SIRT]). Radioembolization is most often used for patients with chemotherapy refractory disease. (See 'Radioembolization' below.)

Combined radioembolization plus systemic chemotherapy has been addressed in patients with chemotherapy-sensitive disease, both in the first-line and second-line settings:

●Chemotherapy-naïve patients – Small randomized trials have suggested potential benefit for combining SIR-Spheres with initial intravenous fluoropyrimidine-based chemotherapy in patients with CRC liver metastases [105,106]. The efficacy and potential for additive toxicity of combined therapy with SIR-Spheres plus modern oxaliplatin-containing conventional chemotherapy over chemotherapy alone was directly addressed in three parallel multinational phase III trials (SIRFLOX, FOXFIRE, and FOXFIRE-Global) in which chemotherapy-naïve patients with CRC liver metastases not amenable to potentially curative resection or ablation were randomly assigned to oxaliplatin-based chemotherapy (with or without an investigator-chosen biologic agent) alone or with a single treatment of radioembolization concurrent with cycle 1 or 2 of chemotherapy [107,108].

In the combined analysis of all three trials (1103 patients, 549 receiving chemotherapy alone and 554 receiving chemotherapy plus radioembolization), at a median follow-up of 43 months, the higher objective response rate with combined therapy (72 versus 63 percent) did not translate into improved median overall survival (22.6 versus 23.3 months, HR 1.04, 95% CI 0.9-1.19) or progression-free survival (11 versus 10.3 months, HR 0.90, 95% CI 0.79-1.02), or a greater likelihood of subsequent liver resection [109]. Furthermore, combined treatment was associated with significantly more grade 3 or 4 adverse events (especially hematologic toxicity). Of the 11 treatment-related deaths on study, eight were in the chemotherapy plus radioembolization group, three of which were attributed to radiation-induced liver disease.

●Use with second-line chemotherapy – The addition of radioembolization to systemic chemotherapy may improve response rates and prolong time to tumor progression for patients receiving second-line systemic chemotherapy, but overall survival is not impacted, and treatment-related toxicity is clearly worse. Additional study is needed to better define the patient population that might benefit from combined therapy before this approach can be considered standard.

In the EPOCH trial, 428 patients who progressed after initial oxaliplatin–based (64 percent) or irinotecan-containing (36 percent) systemic chemotherapy were randomly assigned to second-line chemotherapy (either oxaliplatin- or irinotecan-based, depending of what regimen was used first-line) with or without a single treatment of radioembolization to one or both lobes [110]. Combined therapy was associated with a higher objective response rate (34 versus 21 percent), and modestly longer median progression-free survival (8 versus 7.2 months, HR 0.69, 95% CI 0.54-0.88),but no difference in median overall survival (14.0 versus 14.4 months, HR 1.07, 95% CI 0.86-1.32), and more frequent grade 3 adverse effects (68 versus 49 percent).

CHEMOTHERAPY-REFRACTORY DISEASE

Choice of approach — A nonsurgical local or regional treatment approach for isolated or liver predominant liver metastases is a reasonable option after the cancer has progressed on several systemic therapies; such an approach may improve symptoms and prolong survival. (See "Systemic therapy for nonoperable metastatic colorectal cancer: Approach to later lines of systemic therapy".)

In most cases, the extent of disease will preclude tumor ablation or stereotactic body radiotherapy (SBRT), and transarterial chemoembolization (TACE) and radioembolization are typically the most appropriate options. The choice is usually based on local expertise, patient preference, and disease/patient characteristics. (See 'Transarterial (chemo)embolization' below and 'Radioembolization' below.)

We restrict radioembolization to patients who meet the following criteria:

●Unresectable metastatic hepatic disease with liver-dominant tumor burden, and a life expectancy >3 months.

●An absolute contraindication to yttrium 90 (90Y) microsphere treatment is a pretreatment 99mTc macroaggregated albumin scan that demonstrates the potential for ≥30 Gy radiation exposure to the lung or flow to the gastrointestinal tract that cannot be corrected by catheter techniques.

●Relative contraindications include limited hepatic reserve, irreversibly elevated serum bilirubin levels, compromised portal vein (unless selective or superselective radioembolization can be performed), and prior radiotherapy involving the liver.

Transarterial (chemo)embolization — Transarterial embolization without and with chemotherapy (TACE) has been investigated in patients with chemorefractory CRC liver metastases using both conventional techniques and drug eluting beads. (See "Localized hepatocellular carcinoma: Liver-directed therapies for nonsurgical candidates not eligible for local thermal ablation", section on 'TACE and bland particle embolization'.)

Data from mainly retrospective experience suggests that TACE can achieve disease stabilization in 40 to 60 percent of treated patients, including those with chemorefractory liver-predominant metastatic CRC, with survival durations that may approach a year or more [111-123]. Outcomes seem to be better than can be achieved with continued systemic chemotherapy. As examples:

●Data are available from a single center experience of TACE with 564 patients with isolated liver metastases from CRC who were repeatedly treated with TACE (mean six sessions per patient); 84 were treated in the neoadjuvant setting, the remainder with palliative intent [114]. The median survival after the first TACE procedure was 14.3 months, and one-, two-, and three-year survival rates were 62, 28, and 7 percent, respectively. Notably, survival was significantly better when TACE was performed after first- or second-line systemic chemotherapy (1 to 12 months) than after third- to forth-line therapies (6 months). Evaluation of local tumor control demonstrated a partial response in 17 percent, and prolonged period of stable disease in 48 percent.

●The only data to directly compare TACE versus systemic palliative chemotherapy come from a small phase III trial in which 74 patients with isolated hepatic metastases from CRC were randomly assigned to two monthly infusions of hepatic intra-arterial irinotecan drug eluting beads (DEBIRI) or four months of IV therapy with irinotecan plus short-term infusional fluorouracil and leucovorin (FOLFIRI) [112]. All patients had received at least two or three lines of prior chemotherapy, none in the three months prior to enrollment. The DEBIRI group had a significantly higher objective response rate (69 versus 20 percent), overall survival (56 versus 32 percent at two years, median 22 versus 15 months), and progression-free survival (7 versus 4 months). While the median time to hepatic progression was significantly longer in the group treated with DEBIRI (7 versus 4 months), the median time to extrahepatic progression was also inexplicably longer in this group (median 13 versus 9 months).

Radioembolization — Patients with CRC liver metastases who are unsuitable for resection may benefit from radioembolization. However, given the risks of radiation-induced liver disease, and the tremendous expense of this therapy, it is difficult to know how to incorporate Y-90 radioembolization into the treatment scheme for patients with unresectable hepatic metastases from CRC. We tend to reserve this approach for patients with chemorefractory disease who have a liver-predominant tumor burden and no other therapeutic options.

We agree with the conclusions of a consensus panel from the Radioembolization Brachytherapy Oncology Consortium that suggested that radioembolization be limited to patients who meet the following criteria [124]:

●Unresectable metastatic hepatic disease with liver-dominant tumor burden, and a life expectancy >3 months.

●An absolute contraindication to 90Y microsphere treatment is a pretreatment 99mTc macroaggregated albumin scan that demonstrates the potential for ≥30 Gy radiation exposure to the lung or flow to the gastrointestinal tract that cannot be corrected by catheter techniques.

●Relative contraindications include limited hepatic reserve, irreversibly elevated serum bilirubin levels, compromised portal vein (unless selective or superselective radioembolization can be performed), and prior radiotherapy involving the liver [125].

In contrast to approaches in which the radiation source is external to the patient (ie, external beam radiation therapy, SBRT), alternative means of delivering focal radiation to the liver employ radioactive isotopes (eg, 131-labeled lipiodol or 90Y-tagged glass or resin microspheres) that are delivered selectively to the tumor via the hepatic artery (radioembolization, also called selective internal radiotherapy [SIRT]). After injection into the branch of the hepatic artery that perfuses the region of the liver with the metastasis, the microspheres become preferentially lodged in the arteriolar vasculature surrounding the tumor, delivering high doses of radiation to the area. Maximum tissue penetration for the pure beta-emitter 90Y is 1.1 cm, so most normal liver parenchyma is spared.

SIR-Spheres, a product consisting of 90Y-labeled biocompatible resin microspheres (20 to 40 micrometers in diameter), is available in North America and approved in the United States (concurrent with hepatic intraarterial injection of fluorodeoxyuridine [FUDR]) for the treatment of unresectable liver metastases from primary CRC. Approval was based upon results from a single controlled trial, in which 74 patients with liver-isolated CRC metastases (60 chemotherapy-naïve) were randomly assigned to hepatic intra-arterial chemotherapy with FUDR alone or in conjunction with a single intrahepatic artery administration of SIR-Spheres [126]. Combined therapy was associated with a significantly better objective complete response rate (44 versus 18 percent) and median time to progression (16 versus 10 months), and similar grade 3 and 4 toxicity. Although the one-, two-, three-, and five-year survival rates for patients receiving SIR-Spheres (72, 39, 17, and 4 percent, respectively) did not differ significantly from those of patients in the control arm (68, 29, 7, and 0 percent, respectively), Cox regression analysis suggested a survival benefit for patients who lived longer than 15 months. Notably, at least three adequately powered phase III trials have failed to demonstrate a benefit for radioembolization plus systemic chemotherapy over systemic chemotherapy alone for chemotherapy-naïve disease, and this is generally not considered a standard approach. (See 'Radioembolization plus systemic chemotherapy' above.)

Several studies note benefit from Y-90 radioembolization alone in patients with chemorefractory liver-predominant metastatic CRC [127-134]. A year 2014 systematic review of 20 studies comprising 979 patients found that rates of complete radiologic response, partial response, and stable disease were 0 (range 0 to 6), 31 (range 0 to 73), and 40.5 (range 17 to 78) percent, respectively [135]. Approximately 57 percent had a serologic response (reduction in serum carcinoembryonic antigen). The median time to intrahepatic disease progression was 9 months, median time to disease progression overall was 4.9 months, and the median overall survival was 12 months. Overall, acute toxicity developed in 11 to 100 percent (median 41 percent), most of which was mild (grade 1 or 2). However, the wide range of delivered radioactivity, different treatment volume (ie, lobar versus whole liver) and method/timing of posttreatment evaluation, and variable percentages of patients with extrahepatic disease and concurrent use of chemotherapy in some studies compromise the ability to interpret the results.

Treatment-related toxicity is not trivial. Among the hepatic parenchymal changes that are described in response to radioembolization are transient transaminase elevation and hyperbilirubinemia, ipsilateral hepatic lobar volume decrease with contralateral lobar hypertrophy (a phenomenon that has been termed "radiation lobectomy"), induction of liver fibrosis, and portal hypertension. The frequency and risk factors for development of these adverse effects and their influence on liver insufficiency are not well characterized. Complications of radioembolization are discussed in more detail elsewhere. (See "Localized hepatocellular carcinoma: Liver-directed therapies for nonsurgical candidates not eligible for local thermal ablation", section on 'Complications'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Colorectal cancer".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Colon and rectal cancer (The Basics)")

●Beyond the Basics topics (see "Patient education: Colon and rectal cancer (Beyond the Basics)" and "Patient education: Colorectal cancer treatment; metastatic cancer (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Chemotherapy-naïve disease

•Suitable for resection – Surgical resection is the treatment of choice for patients with isolated, potentially resectable colorectal cancer (CRC) liver metastases, when feasible. Some patients may be offered initial systemic chemotherapy with deferred resection. Specific recommendations are provided separately. (See "Potentially resectable colorectal cancer liver metastases: Integration of surgery and chemotherapy", section on 'Patients with initially unresectable metastases'.)

•Not suitable for resection – Patients with isolated liver metastases may be deemed unsuitable for resection because of tumor location, impaired general health status, or an insufficient future liver remnant to resect all lesions. For such patients who have a limited number of small lesions, we suggest locoregional liver-directed treatment using radiofrequency ablation (RFA), microwave ablation (MWA), or stereotactic body radiotherapy (SBRT), which is an acceptable alternative to initial systemic chemotherapy. One advantage of this approach is deferral of chemotherapy-related toxicity. However, palliative systemic chemotherapy is an option, particularly if local expertise to perform nonsurgical regional therapy is not available, and patients have not received prior adjuvant chemotherapy. One commonly employed strategy is to start with chemotherapy, and consolidate this, for appropriate candidates, with subsequent ablation. (See "Systemic therapy for nonoperable metastatic colorectal cancer: Selecting the initial therapeutic approach" and "Therapy for metastatic colorectal cancer in older adult patients and those with a poor performance status".)

The choice of specific locoregional treatment, tumor ablation with RFA or MWA, or SBRT, usually depends on local expertise and patient preference, as well as the number, size, and location of the lesions.

Patients who are ineligible for locoregional therapy because of disease extent, or poor underlying liver function, or if locoregional therapy is not available are treated with systemic chemotherapy. Specific recommendations are provided separately (see "Systemic therapy for nonoperable metastatic colorectal cancer: Selecting the initial therapeutic approach" and "Therapy for metastatic colorectal cancer in older adult patients and those with a poor performance status").

The benefit of ablation (RFA, MWA, or SBRT) for regional treatment of liver metastases who are undergoing systemic chemotherapy for chemotherapy-sensitive metastatic CRC remains unproven, although it is an option for patients with a limited burden of liver-limited disease following initial systemic therapy. We suggest against adding radioembolization to initial systemic chemotherapy given that combined therapy does not improve survival but does increase adverse events (Grade 2B). (See 'Local ablation plus systemic chemotherapy' above.)

●Chemotherapy refractory-disease

•A nonsurgical locoregional liver-directed treatment approach is a reasonable option for patients with isolated or liver-predominant liver metastases and adequate underlying liver function after the cancer has progressed on several systemic therapies. (See 'Chemotherapy-refractory disease' above.)

•In most cases, the extent of disease will preclude tumor ablation or SBRT, and transarterial chemoembolization or radioembolization are appropriate options. The choice is usually based on local expertise, patient preference, and patient/tumor characteristics. (See 'Transarterial (chemo)embolization' above and 'Radioembolization' above.)