Locoregional Therapy for Intrahepatic Cholangiocarcinoma

Simple Summary Intrahepatic cholangiocarcinoma is an aggressive primary liver cancer originating in the intrahepatic bile ducts. While surgical resection is the only curative treatment, many patients present with locally advanced, unresectable, or metastatic disease, and few are candidates for curative-intent resection. In this review, we examine locoregional therapy approaches and summarize the current literature. Current locoregional therapies include thermal ablation, transarterial chemoembolization, transarterial radioembolization, external beam radiotherapy, stereotactic body radiotherapy, hepatic arterial infusion of chemotherapy, irreversible electroporation, and brachytherapy. These therapies are most often offered to patients with unresectable primary or recurrent intrahepatic cholangiocarcinoma, and studies on each modality have shown these locoregional approaches to be effective for prolonging overall survival. The findings of this review also further inform the need for future research regarding the efficacy of these treatments in comparison to each other due to the limited literature on optimal treatment strategies. Abstract Intrahepatic cholangiocarcinoma (ICC) has a poor prognosis, and surgical resection (SR) offers the only potential for cure. Unfortunately, only a small proportion of patients are eligible for resection due to locally advanced or metastatic disease. Locoregional therapies (LRT) are often used in unresectable liver-only or liver-dominant ICC. This review explores the role of these therapies in the treatment of ICC, including radiofrequency ablation (RFA), microwave ablation (MWA), transarterial chemoembolization (TACE), transarterial radioembolization (TARE), external beam radiotherapy (EBRT), stereotactic body radiotherapy (SBRT), hepatic arterial infusion (HAI) of chemotherapy, irreversible electroporation (IE), and brachytherapy. A search of the current literature was performed to examine types of LRT currently used in the treatment of ICC. We examined patient selection, technique, and outcomes of each type. Overall, LRTs are well-tolerated in the treatment of ICC and are effective in improving overall survival (OS) in this patient population. Further studies are needed to reduce bias from heterogenous patient populations and small sample sizes, as well as to determine whether certain LRTs are superior to others and to examine optimal treatment selection.


Introduction
Intrahepatic cholangiocarcinoma (ICC) is a rare and aggressive primary hepatic malignancy with increasing incidence in the United States (US) and worldwide. Between 1973 and 2012, the incidence of ICC has risen from 0.4 to 1.18 cases per 100,000 persons in the USA and represents the second most common primary liver cancer behind hepatocellular carcinoma (HCC) [1]. ICC carries a poor prognosis with a 5-year OS of less than 10% and increasing mortality rates [1,2]. The only curative treatment for ICC is surgical resection (SR), though only up to 30% of patients are eligible because the disease is often locally advanced or metastatic at the time of presentation [3]. Even with SR, the 5-year OS remains low at 22-45%, and recurrence rates are as high as 80% [4]. For patients who are able to undergo SR, adjuvant capecitabine is recommended based on the results of the BILCAP trial, which demonstrated improved OS for adjuvant capecitabine in the per-protocol analysis [5]. Locoregional therapies (LRTs) have also been compared to SR for the primary treatment of early-stage disease or used in the adjuvant setting, after SR.
For patients with unresectable or metastatic disease, clinical trials, systemic therapy, chemoradiation (ChR), and LRTs are among the treatment options [6]. On the basis of the ABC-02 trial, gemcitabine and cisplatin doublet therapy are the preferred regimens for systemic therapy in the first-line setting [7]. Triplet therapy with durvalumab, gemcitabine, and cisplatin was also recently adopted as a preferred option in the first-line setting after demonstrating improved OS compared to placebo plus chemotherapy in the TOPAZ-1 study [8]. FOLFOX is the preferred regimen in the second-line setting for patients who progress on gemcitabine and cisplatin [9].
For patients with unresectable ICC with a liver-only or liver-predominant disease, LRTs represent a promising treatment option for multimodality treatment [6,10]. LRTs currently in use for ICC include RFA, MWA, TACE, TARE, EBRT, SBRT, HAI, IE, and brachytherapy. Preliminarily, these LRTs have shown promising results in the treatment of ICC, with improved survival rates [11]. The objective of this review is to discuss the LRTs currently in use and to summarize the current literature.

Methods
In this narrative review, we summarized articles related to locoregional therapies for intrahepatic cholangiocarcinoma. We included articles published between 2005 and 2022. We included articles with data on outcomes from the following locoregional therapies: RFA, MWA, TACE, TARE, EBRT, SBRT, HAI, IE, and brachytherapy. Articles were excluded if they focused on extrahepatic cholangiocarcinoma including distal or hilar cholangiocarcinoma, or gallbladder cancer, or if they included patients with all biliary tract cancers and the data for patients with intrahepatic cholangiocarcinoma could not be separated.

Patient Selection
RFA is an LRT that can treat a variety of solid tumors, with predominant use in the liver. For ICC specifically, RFA represents a promising option for patients who are not candidates for curative SR due to advanced cancer at diagnosis, poor hepatic reserve, or serious comorbidities [12,13]. RFA also represents a treatment option for patients who have recurrence after SR [14,15]. In examined studies, exclusion criteria for RFA in ICC varied, though generally patients with severe coagulopathy, severe thrombocytopenia, vascular invasion, tumor size > 5-7 cm, multiple hepatic lesions > 3-5, progressive extrahepatic metastases, or poor performance status were excluded [13,[16][17][18].

Technique
RFA is a minimally-invasive technique that is most commonly performed percutaneously by an interventional radiologist with the patient under general anesthesia utilizing imaging-guidance including ultrasound (US) and computed tomography (CT) [18,19]. Procedural characteristics, as well as outcomes, are summarized in Table 1. RFA can also be performed in an open fashion, intraoperatively, and/or in conjunction with SR [20]. The technique utilizes a number of needle electrodes, depending on tumor size and location, often set at a 200 W current for a range of 10-90 min to ablate visible tumors with margins of 5-10 mm [16,21]. The high frequency of the electric current emitted by the electrode generates frictional heat, which causes localized cell death [22]. Technical success, defined as the treatment of the tumor according to protocol, has been reported at rates ranging from 80-100% [23]. Patients are often followed with multiphasic CT or magnetic resonance imaging (MRI) to assess for imaging response, typically obtained 1 month after the procedure [23]. Technical effectiveness based on complete imaging response at 1 month similarly ranged from 80-100% [21]. Lower rates of effectiveness were observed in patients with larger tumors, often >5 cm [12,24]. Local recurrences were often treated with repeat RFA [16,19,25,26]. Abbreviations: ChR = chemoradiation; CT = computed tomography; DFS = disease-free survival; ICC = intrahepatic cholangiocarcinoma; IORFA = intraoperative radiofrequency ablation; LTPFS = local tumor progression-free survival; mo = month; MWA = microwave ablation; ND = no difference between groups; OS = overall survival; RFA = radiofrequency ablation; RFT = recurrence-free time; S = number of sessions; SFRA = stereotactic radiofrequency ablation; SR = surgical resection; TTR1 = time to 1st recurrence relative to primary surgical resection; TTR2 = time to 2nd recurrence relative to primary surgical resection; US = ultrasound; x = occurrences; yr = years; y/o = years old.
The studies examined in this review contained varying patient populations, including those with early-stage ICC to those with recurrent, unresectable ICC. For patients with unresectable or recurrent ICC treated with RFA, median OS ranged from 20-60 months, compared to median OS rates of 3-8 months in patients with unresectable ICC who did not undergo any treatment [6,[12][13][14][15]17,18,20,[26][27][28][29]32]. Interestingly, Xiang et al. found that SR showed significantly improved 1-, 3-and 5-year OS and cancer-specific survival (CSS) rates compared to RFA for stage I tumors < 5 cm [17]. On the other hand, Wu et al. found that RFA conferred a significant 5-year OS benefit for stage I tumors < 5 cm compared to ChR with an OS rate of 20.1% compared to 3.7%, respectively [32]. Overall, these results are encouraging, though further randomized controlled trials with larger sample sizes and prospective designs are warranted.

Technique
MWA is a minimally-invasive technique most commonly performed with patients under general anesthesia while a microwave probe is inserted percutaneously into the tumor under imaging-guidance [33]. MWA emits electromagnetic radiation to induce tumor cell death via frictional heat, though is capable of generating higher temperatures in shorter time periods, with the goal of more completely ablating targeted tissue and avoiding nearby structures [22]. In contrast to RFA, MWA utilizes uniform heating, has more predictable ablation zones that can target larger liver volumes, and can treat multiple lesions simultaneously, making it a frequently used modality [45]. Included studies used a variety of settings for MWA. Most commonly, the output power was set to 40-100 W for 3-20 min [25,33,34,36,37,39,41,42]. Single electrodes were often used for tumors < 2-3 cm, while multiple electrodes or ablations were used for larger tumors [25,33,34,37,[41][42][43][44]. Tumors were treated with the goal of achieving 0.5-1 cm margins, and the needle tract was ablated during the removal of the probe to avoid tumor seeding [25,26,34,38,39,42,43]. Technical success was defined as the ability to treat the tumor according to protocol and was assessed 2-5 days after treatment with contrast-enhanced (CE) CT or MRI [26,33,34,39,43,44]. Technical effectiveness was defined as complete ablation 1 month after the procedure, and most studies describe subsequent ablations if a residual tumor was present [36,38,40,43,44]. After an initial 1-month follow-up, patients generally followed up every 3-6 months [25,33,34,[36][37][38][39][40][41][42][43].
Survival data varied widely between studies, likely due to the variable patient populations included in each study. Of studies examined in this review, the median OS of ICC patients treated with MWA ranged from 8.8-31.5 months, and median progressionfree survival (PFS) ranged from 6.2-18.43 months [25,26,33,35,36,[40][41][42][43][44]. Interestingly, the study by Yan et al. assessed the impact of combining thermal ablative therapy, consisting of RFA or MWA, with systemic chemotherapy, compared to chemotherapy alone in the treatment of unresectable and previously untreated ICC. They found that the median OS was significantly higher in the combined group (combined median OS = 15.23 months vs. chemotherapy alone = 7.97 months, p = 0.009) [34]. Additionally, the study by Giorgio et al. was unique in that it directly assessed OS and PFS in patients with unresectable ICC treated with RFA vs. MWA [38]. This study demonstrated a statistically significant increase in both OS and PFS favoring MWA [38]. Finally, the study by Xu et al. was notable for assessing median OS and complication rates in patients with recurrent ICC after initial SR, treated with MWA vs. repeat SR [40]. They found no significant difference in median OS between the two groups, but the repeat SR group had a statistically significant increase in major complications (MWA = 5.3% vs. repeat SR = 13.8%, p < 0.001) [40]. While data between studies are heterogeneous, results show that MWA is both well-tolerated and similarly effective in comparison to other LRTs.

Patient Selection
TACE is an intra-arterial therapy that has been employed in various hepatic malignancies, including ICC. While TACE is a non-curative therapy, it represents an option for locoregional tumor control in patients ineligible for SR, as an adjuvant therapy after SR, or in patients with progression of disease after initial treatment [46,47]. Inclusion and exclusion criteria varied between studies, and eligible patients generally required adequate performance status as well as sufficient bone marrow, liver, and renal function [48,49].
Patients with severe comorbidities, active infection, or contraindication to arterial procedures were generally excluded [48,49]. Tumor size was not mentioned as an exclusion criterion in the examined studies for TACE. This is in contrast to RFA, where patients with tumors >5 cm were often considered ineligible.

Technique
TACE represents a treatment option for patients with locally advanced tumors, with the goals of delivering higher concentrations of chemotherapeutic agents locally and engendering tumoral ischemia [50]. The technique, outcomes, and complications for TACE in included studies are summarized in Table 3A. During the procedure, the hepatic artery supplying the tumor is most commonly accessed via a femoral approach and identified with conventional angiography [48,49,51]. In conventional TACE (cTACE), a chemotherapeutic agent emulsified with lipiodol is injected into the hepatic artery followed by arterial occlusion with embolic material, depriving the tumor of blood supply [51]. Commonly used chemotherapies include doxorubicin, cisplatin or carboplatin, mitomycin-C, and gemcitabine [46]. In a slightly different procedure, drug-eluting bead TACE (DEB-TACE) employs chemotherapy-laden beads or microspheres to both deliver the medication and embolize the artery simultaneously [49,52]. In both cTACE and DEB-TACE, the procedure is complete when near-stasis is achieved on angiography. Most studies assessed tumor response to treatment with follow-up imaging using CT or MRI according to the modified Response Evaluation Criteria in Solid Tumors (mRECIST) criteria [48,49,[52][53][54]. Repeat sessions of TACE can be performed for residual tumors, recurrence, or progression of the disease.

Outcomes
Similar to RFA, studies have shown that TACE is a safe and well-tolerated treatment method for ICC. Major complications were rare but included inguinal hematoma, hepatic arterial dissection, hepatorenal syndrome, severe thrombocytopenia, and hepatic abscess [53,55,56]. Major complications occurred at a rate of 0-12.5% [3,49,52,53,55,57,61]. Minor complications included post-embolization syndrome, which consists of abdominal pain, fatigue, nausea, vomiting, and fever, as well as transient decreases in liver function [61]. Luo [48]. While these results are promising, again, larger studies and prospective designs are warranted to further assess the efficacy of TACE as an LRT option for ICC.

Technique
The technique utilized in TARE is similar to that of TACE. Prior to treatment, the hepatic vasculature and pulmonary shunt fraction is assessed via a planning session including diagnostic angiography and administration of 99mTc macroaggregated albumin to the targeted tumoral arterial distribution [62,[64][65][66][67][68][69]71,72,75,76,78]. Following pre-treatment mapping, target vessels are injected with yttrium-90 (Y90) resin or glass microspheres in the treatment session, delivering localized radiation doses to the tumor, while sparing nearby normal tissue [62,76,77]. The median administered Y90 activity ranged from 1.5-1.74 GBq in included studies [62,64,65,67,68,73,74,77,78]. Patients were treated with multiple sessions as needed [62,63,65]. Duration of hospital stay after the procedure varied considerably by study, with some protocols discharging patients 2-4 h after the procedure and others discharging patients after 1-4 days [62,70,72,76,78]. Most patients were assessed at 1 month for a response to treatment using the RECIST criteria and CT, MRI, or positron emission tomography (PET) imaging, followed by visits every 3 months [ Outcomes for TARE varied widely by study, likely due to small sample sizes and variable patient populations. Median OS ranged from 5.7-33.6 months and median PFS ranged from 2.8-10.1 months [62][63][64][65][66][67][68][69][70][71][72][74][75][76][77][78]. The study by Bargellini et al. was notable for assessing the impact of TARE in three groups: Group A-chemotherapy naïve patients treated with TARE, Group B-patients treated with chemotherapy then adjuvant TARE, and Group C-chemotherapy-refractory patients treated with TARE [64]. The study found no significant difference in OS between the groups and the median OS was 14.5 months [64]. Interestingly, the study by Buettner et al. assessed the impact of using resin versus glass microspheres for TARE, though found no significant difference in median OS (29 months) and median PFS (5 months) between the groups [65]. Finally, the case series by Filippi et al. was unique in assessing the impact of repeat TARE in patients with recurrent ICC after the first TARE [67]. The mean time between the first TARE and recurrence was 7.3 months and the median OS after the first TARE was 16.5 months, though the median OS after the second TARE was unfortunately not reported [67].

Patient Selection
EBRT is an LRT used in the treatment of many cancers and is indicated in ICC as well. Like other LRTs, EBRT is not a curative treatment, though can be used in cases of unresectable ICC, recurrent ICC, as an adjuvant therapy after SR, in combination with chemotherapy, or as a palliative treatment [79,80]. Examined studies are summarized in Table 5. In these studies, few patients were considered ineligible for EBRT. Exclusion criteria consisted of patients with Child-Pugh class C cirrhosis, other primary liver tumors, or other serious conditions [81][82][83]. Tumor size was not a factor in determining eligibility and patients with tumor sizes ranging from 2.2-17 cm were treated in the included studies [84].

Technique
Several different techniques are available for the delivery of EBRT. Studies examined in this review delivered radiation using a linear accelerator with 6-Megavolt (MV) or 15 MV photons or via passive scatter photon beam techniques [82][83][84]88]. The recent study by Smart et al. was unique in that it used hypofractionated photon or proton beams [81]. Before treatment, the size and location of the treatment field were determined by 2-dimensional (2D) or 3-dimensional (3D) CT or MRI imaging [81][82][83][84]88]. Radiation planning also required the determination of gross tumor volume (GTV), clinical target volume (CTV), and planning target volume (PTV). These measures were variably defined by each study and included margins to account for setup error and respiratory movement [81][82][83][84]88]. Median overall radiation dose ranged from 50-58.05 Gray (Gy), most often delivered in daily 2 Gy fractions 5 times a week [81][82][83][84]88]. Notably, Smart et al. treated patients with 15 Gy daily fractions to a median overall dose of 58.05 Gy [81]. Patients were monitored clinically every week during treatment. Response to EBRT was assessed initially at 6 weeks via CT or MRI, followed by monitoring every 3 months [81][82][83]88].

Outcomes
Studies have demonstrated that EBRT is well tolerated. In examined studies, Grade 3 or higher complications occurred at a rate of 12.5% or less [81][82][83][84]88]. Commonly reported adverse effects included neutropenia, thrombocytopenia, elevations in LFTs, anorexia, nausea, vomiting, abdominal pain, fatigue, and fever. Chen et al. reported one case of RILD that resulted in mortality [82]. Smart et al. reported one case of RILD that was treated with glucocorticoids as well as cases of ascites, GI bleeding, hepatomegaly, and the need for biliary intervention after treatment [81]. Tao et al. noted that 5 patients were hospitalized for complications related to EBRT or tumor progression [84]. The need for hospitalization was not mentioned in other studies. Overall, EBRT is safe and well tolerated.
Response to treatment and OS varied significantly between studies because included patient populations and study designs were diverse. Median OS ranged widely from 7-39.5 months [81,82,[84][85][86]88,89]. Notably, the retrospective study by Kolarich et al. using the National Cancer Database (NCDB) compared the use of EBRT, RFA, radioactive implants (RI), and no local treatment in nonsurgical ICC patients with stage I-IV disease, and found statistically significant improved median OS for patients with stage I disease receiving EBRT or RFA, stage III disease receiving EBRT or RI, and stage IV disease receiving RI compared to no local therapy [85]. Interestingly, the retrospective cohort study by Hammad et al. using the NCDB examined the impact of adjuvant EBRT after SR and found statistically significant improved OS for patients with R0 resection compared to R1/R2 resections (31.2 vs. 19.5 months, p < 0.001); however, after multivariate analysis, adjuvant EBRT was not associated with survival in patients with R1/R2, lymph-nodenegative resections [86]. On the other hand, Jackson et al., in another retrospective study using the NCDB, demonstrated improved OS for patients with unresectable, localized ICC receiving EBRT and chemotherapy compared to chemotherapy alone (2-year OS 25.8% vs. 20%, p = 0.001) [87]. Finally, Shao et al., in a retrospective study using The Surveillance, Epidemiology, and End Results database (SEER), showed improved OS (p = 0.00228) and cancer-specific survival (CSS) (p = 0.0037) in patients receiving palliative EBRT compared to no EBRT [80]. As a whole, while results were heterogeneous, these studies indicate that EBRT is an effective treatment for ICC in specific patient populations and especially compared to patients not receiving any LRT.

Patient Selection
Patient selection for use of SBRT and EBRT in the treatment of ICC is similar. Studies included in this review examined the use of SBRT for patients who had previously received SR with positive surgical margins, as adjuvant therapy after SR, for primary unresectable tumors, locally recurrent disease, or after other previous LRT or chemotherapy (Table 6) [90][91][92][93]. Patients included in these studies required adequate performance status [94]. Exclusion criteria varied by study though consisted of patients with <3-4 lesions, other serious medical conditions, other malignancies, inadequate liver function or volume, inadequate bone marrow function, or prior abdominal radiation [90,[94][95][96]. The retrospective study by Sandler et al. limited selection to patients with tumors < 8 cm, though other studies included tumors > 10 cm [94,95,[97][98][99][100].

Outcomes
Similar to EBRT, SBRT is well-tolerated in the treatment of ICC. In included studies, Grade 3 complications ranged from 4.7-53% and Grade 4-5 complications ranged from 0-11% [91,92,[94][95][96][97][98][99][100]. The most commonly reported adverse effects were fatigue, nausea, abdominal pain, anorexia, bone marrow suppression, and elevated liver enzymes [91,92,[94][95][96][97]100]. More severe complications were rare and included GI bleeding, bowel obstruction, biliary obstruction or stenosis, acute cholecystitis, acute cholangitis, liver abscess, a decline in Child-Pugh class, and liver failure [91,92,[94][95][96][97]100]. Interestingly, the rates of Grade 3-5 adverse effects for SBRT in included studies were higher than those for EBRT. This may indicate that despite precise targeting in SBRT, higher radiation doses can still cause complications. However, the interpretation of these data may be complicated by the fact that complications were less frequently reported in included studies for EBRT. Further studies are needed to directly evaluate complication rates for EBRT versus SBRT in the treatment of ICC. Regardless, both are well-tolerated options for LRT.
Outcome data surrounding the use of SBRT for ICC are difficult to interpret due to varied study designs and patient populations. Of examined studies, the median OS ranged from 13.2-23 months [91,[94][95][96][97][98][99][100]. PFS ranged from 6.1-24.7 months [91,92,[94][95][96]99,100]. These data are challenging to interpret because many of the examined studies included both patients with ICC and extrahepatic cholangiocarcinoma (ECC) and did not separate outcome data between the two diseases [91,92,95,98,99]. This is problematic because studies have shown that these cancers may represent distinct disease processes with different epidemiology, risk factors, genetic makeup, recurrence trends, and response to treatment [103][104][105]. Notably, studies by Kozak et al., Shen et al., Weiner et al., and Tse et al. separated outcome data for ICC specifically and examined the use of SBRT as adjuvant therapy after SR, definitive treatment and for unresectable disease [94,96,97,100]. Finally, the retrospective cohort study by Sebastian et al. was interesting in that it compared the use of SBRT, ChR, and TARE for unresectable ICC [101]. This study found that OS was significantly greater for patients treated with SBRT compared to ChR (p < 0.0001) and identified no significant difference in OS between SBRT and TARE [101]. As with other LRTs, more data are needed to examine the impact of SBRT in the treatment of ICC and in comparison to other treatment modalities.
As with studies on other LRTs for ICC, response to treatment varied by study, likely due to the small sample size and variable patient population. Median OS ranged from 10.1-31.1 months and median PFS ranged from 5-11.8 months [106][107][108][110][111][112][113][114][115][116][117][118][119][120][121]. The study by Franssen et al. was notable for assessing outcomes among patients with multifocal ICC who received either HAI or SR and found no significant difference in median OS, though SR was associated with a significant increase in Grade 3A adverse effects (HAI = 6.4% vs. SR = 25.3%, p = 0.04) [106]. Additionally, the study by Ishii et al. found a statistically significant increase in median OS (19.7 months vs. 10.8 months, p = 0.006) for patients with advanced ICC receiving HAI vs. systemic chemotherapy [108]. Again, more data are needed to further assess the impact of HAI, especially in comparison to other LRTs.

Irreversible Electroporation
While several studies have examined the use of IE in the treatment of hilar cholangiocarcinoma, data surrounding the use of IE for ICC are extremely limited [123][124][125][126]. Most notably, a prospective study by Belfiore et al. examined the use of IE for unresectable ICC (N = 8) and perihilar (PHCC) cholangiocarcinoma (N = 7) between 2015-2019. In this study, exclusion criteria consisted of cardiopulmonary failure, inadequate hematologic and bone marrow function, contraindications to general anesthesia, bilirubin >3 mL/dL, extrahepatic metastases, recent myocardial infarction, or active infection. Eligible patients underwent pre-treatment staging with CT, endoscopic retrograde cholangiopancreatography (ERCP), and magnetic resonance cholangiopancreatography (MRCP). The procedure was performed under general anesthesia, and tumors were treated with 2-4 bipolar needles or electrodes that delivered 90 pulses at 1500 V/cm, thus ablating the tumor via non-thermal electrical energy. This technique avoids thermal damage to nearby tissues, as seen in other forms of LRT [127]. Response to treatment was assessed in 1-, 3-, and 6-month intervals after the procedure with multidetector CE-CT and MRCP. Belfiore et al. reported no severe complications. The study did not separate ICC and PHCC outcome data; though, via Kaplan-Meier analysis, the mean OS for the entire group was estimated to be 18 mo [127]. In addition to this study, a randomized clinical trial by Zhang et al. assessed the use of IE and cryoablation in ICC and HCC patients, though the trial was designed to assess the impact of allogeneic gamma delta T-cell transfer on outcomes, not IE efficacy [128]. Finally, Eisele et al. included two ICC patients, among several other liver cancer patients, in the use of US-guided IE therapy. The study only assessed local failure rates (21%), not survival outcomes [129]. More studies are clearly needed to assess the use of IE in ICC.

Brachytherapy
Like IE, the literature on the use of brachytherapy for ICC is sparse. One retrospective study by Schnapauff et al. assessed the use of brachytherapy in 15 patients with unresectable ICC without extrahepatic disease between 2006-2009 [130]. Patients with elevated bilirubin >2.5 mL/dL, >5 liver lesions, impaired coagulability and large-volume ascites were excluded from the study [130]. Patients who were eligible underwent treatment planning with contrast liver MRI. Brachytherapy catheters were percutaneously inserted into tumors under CT guidance, with a target CTV of 20 Gy and a permissible dose of >50 Gy in the central parts of the tumor [130]. Brachytherapy delivers high radiation doses to malignant tissue while sparing normal tissue nearby [131]. Unlike thermal ablative options such as RFA and MWA, this treatment has the potential to be effective in tumors with large diameters >5 cm [130]. Patients were followed regularly with tumor markers and liver MRI at 6-12 weeks, followed by 3-month intervals to assess response to treatment [130]. Repeat treatments with brachytherapy were performed in patients with local tumor progression [130]. Tumor size in the study by Schnapauff et al. ranged from 1-18 cm [130]. The study reported a median control time of 10 months after treatment, a median of 13 months to systemic progression and a median OS of 14 months [130]. No severe complications were reported as a result of brachytherapy, though authors acknowledged that the study population size was small [130]. Similar to IE, more studies are needed to determine the feasibility of brachytherapy for the treatment of ICC.

Conclusions
This review examined the current literature surrounding the use of LRTs for the management of ICC, including RFA, MWA, TACE, TARE, EBRT, SBRT, HAI, IE, and brachytherapy. The majority of studies included in this review were retrospective studies that assessed LRTs for patients with primary unresectable, advanced, or recurrent ICC [66]. Several TACE studies also assessed its use in the adjuvant setting after SR [31,[58][59][60]. Across all treatment modalities, patients were generally considered eligible for LRT if they had adequate performance status, sufficient liver, kidney, hematologic, and bone marrow function, and were without significant extrahepatic metastases, comorbidities, and other cancers. RFA and MWA were two treatment modalities where the size of the tumor, often >3-5 cm, was also an excluding factor [22]. The technique used and mechanism of tumor cell death varied by treatment type, though generally LRTs used thermal energy, electric current, chemotherapeutic agents, or radiation delivered locally under image guidance via insertion directly into the tumor in ablative therapies, hepatic circulation in catheterdirected therapies, or externally as in the case of EBRT and SBRT (Tables 1-7).
LRTs are generally well-tolerated in the treatment of ICC and major complications were rare (Tables 1-7). Common minor complications included fatigue, nausea, vomiting, diarrhea, abdominal pain, anorexia, and hematologic lab abnormalities. The response to treatment and patient outcomes were variable between LRTs and even among different studies examining the same LRT modality, likely due to heterogeneous patient populations and small sample sizes. Overall, LRTs drastically improve overall survival compared to the natural history of ICC without treatment [6]. Several studies also showed LRTs to be superior to systemic chemotherapy alone [34,57,87,108,111,116]. In regard to the efficacy of LRTs in comparison to each other, few studies assessed this question, and limited conclusions can be made, again due to variable patient population and small sample size [3,37,38,73,101,109,110]. Current National Comprehensive Cancer Network (NCCN) guidelines reflect the importance of pursuing SR for eligible patients and consider LRTs to be suitable options for patients with unresectable or metastatic ICC when part of a clinical trial or at an experienced center [132]. Overall, LRTs are safe and effective options for the treatment of unresectable or recurrent ICC, though more research is needed to assess the superiority of LRT options in comparison to each other.

Conflicts of Interest:
The authors declare no conflict of interest.