Gastrointestinal Intervention 2017; 6(2): 94-104
Published online July 31, 2017 https://doi.org/10.18528/gii1500342
Copyright © International Journal of Gastrointestinal Intervention.
Charilaos Papafragkakis, and Jeffrey H. Lee*
Department of Gastroenterology, Hepatology and Nutrition, The University of Texas MD Anderson Cancer Center, Houston, TX, USA
Correspondence to:Department of Gastroenterology, Hepatology and Nutrition, The University of Texas MD Anderson Cancer Center, Unit 1466, 1515 Holcombe Blvd., Houston, TX 77030, USA.
This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Cholangioarcinoma is a rare but dreadful malignancy which poses much difficulties in the management. If detected early with only localized disease, curative resection is possible. However, most patients present in the late stages of the disease, which are managed with endoscopic biliary drainage and/or chemoradiation. Liver transplantation offers a possibility for cure in the distal and the perihilar tumors for selected candidates. Local treatments, such as hepatic artery-based therapies, brachytherapy, and photodynamic therapy, may offer some benefit in cases of the advanced disease. In this review, we will assess the role of preoperative biliary drainage, how best to drain biliary obstruction, and the intricate details of various treatments that are currently available.
Keywords: Cholangiocarcinoma, Cholestasis, Endoscopic ultrasonography, Liver transplantation, Self expandable metal stents
Preoperative biliary drainage (PBD) is achieved with endoscopic or percutaneous approach. The topic is controversial and the review of the literature revealed conflicting results. PBD of the future remnant hepatic lobe is supposed to decrease hepatic dysfunction and liver failure postoperatively.1 The meta-analysis found no difference in mortality between patients with and without PBD and the authors advised that such procedure should not be performed routinely.2 Another meta-analysis of almost 5,000 patients with distal obstruction (425 with distal cholangiocarcinoma [DCC]) did not find any evidence that PBD (without distinguishing between endoscopic or percutaneous approach) increases morbidity or mortality. However it was associated with significant bacterial infection of the bile and possibly adverse outcomes after surgery.3 Other studies showed that PBD reduced jaundice, which positively affected outcomes, but not survival, and that it might be associated with more intraoperative blood loss, but less reoperation rates compared to non-drained cases.4,5 The prospective randomized study of almost 200 patients with cancer of the head of the pancreas demonstrated that PBD was associated with more non-surgical adverse events compared to those managed only with surgery (46% vs 2%, respectively). Surgical adverse events were also more in the PBD group (47% vs 37%). The high incidence of cholangitis in those treated with PBD may be associated with the long lag time between PBD and surgery (5 weeks) and the high occlusion rate of plastic stents (PSs) (15%), prompting other authors to suggest that use of short, self-expandable metal stents (SEMS) may result in better outcomes for distal bile duct strictures. In accordance with other studies, there was no effect of PBD on mortality.6,7 In perihilar cholangiocarcinoma (PCC), the systematic review of 700 patients with and without PBD suggested that there is higher incidence of postoperative infections (18%–52% vs 0%–27%) and overall adverse events (36%–100% vs 29% 72%) in the PBD compared to the non-PBD group. There was no difference in mortality in the two groups.8 Other studies favor the use of PBD in the preoperative management of PCC. The study of 350 patients demonstrated reduction in in-hospital mortality in those treated with endoscopic biliary drainage (EBD) or percutaneous biliary drainage (PTBD).9 Results from a multicenter European retrospective study demonstrated no effect of EBD or PTBD on overall post-surgical mortality; however, there was a suggestion that mortality was higher in patients with PBD and with left compared to right hepatectomy, likely due to more cholangitis episodes in the left hepatectomy group.10
Other studies assessed the use of endoscopic nasobiliary drainage (ENBD) in patients with PCC. The study of 116 patients showed that ENBD was associated with a low rate of pancreatitis and cholangitis.11 Another study compared EBD, ENBD, and PTBD in 128 patients and concluded that all three approaches did not differ significantly in morbidity and mortality.12 However, a higher incidence of cholangitis was observed with EBD (60% vs 10% with ENBD and 2% with PTBD). Nevertheless, PTBD had 6% cancer dissemination and 8% portal vein (PV) injury rates.12 PBD has been shown to benefit patients with < 30% functional liver remnant (FLR) as shown in a study of 60 patients that were managed with PTBD and EBD. The absence of PBD in patients with FLR < 30% was associated with a higher incidence of liver dysfunction (as measured by persistent rise in bilirubin) and death.13
In conclusion, it seems that for DCC, PBD has to be applied on a case-by-case basis. If surgery is planned within 1 to 2 weeks, PBD is usually not recommended. If surgery is delayed more than 3 weeks, then PBD should be considered.14 Strong debate still continues whether PBD should be used in proximal biliary cancers. However, PBD would provide benefits in cases such as usage as a bridge to surgery, treatment of cholangitis, correction of malnutrition, correction of coagulopathy, treatment of jaundice-induced liver or kidney dysfunction, induction of hypertrophy of the post-surgical liver and plan for neo-adjuvant therapy. Nonetheless, potential disadvantages of PBD such as tumor seeding, injury of vasculature, perforation and increased difficulties during surgery (e.g., in bilioenteral anastomosis after natural downsizing of the bile duct following stent removal) should be also considered (Table 1).12,15–17
Occasionally, endoscopic retrograde cholangiopancreatography (ERCP) is unsuccessful and biliary drainage remains to be achieved by PTBD or surgery. A new approach with endoscopic ultrasound (EUS)-guided biliary drainage (EGBD) provides a potentially useful alternative. With this technique, a fistula is created between the duct proximal to the obstruction and the duodenum or the stomach. Currently there are no studies with regard to EGBD specifically in the setting of PBD.18 The recent study of EGBD in 101 patients showed that the procedure had very high success rate (> 92% overall) and there was no significant difference in efficacy among different approaches.19 There was, however, 12% rate of adverse events, mostly in the hepatogastrostomy approach, and notable procedure-associated mortality (6 deaths).19 The technical success of EGBD was found to be inferior to the PTBD (86% vs 100%, respectively) for distal malignant biliary strictures. PTBD, however, was associated with more adverse events, such as bile leak and hemobilia and increased need for re-intervention.20 Another multicenter retrospective study in 240 patients showed that EGBD had high success rate in both intra- and extrahepatic approach, but was associated with a high incidence of adverse events, mainly bleeding (11%) and bile leak (10%).21
Endoscopic biliary sphincterotomy is not required prior to PS or SEMS insertions and this practice is not associated with decreased stent patency or increased risk of pancreatitis.22,23 According to the 2015 American Society of Gastrointestinal Endoscopy guidelines, antibiotics to cover gram negatives and enterococci should be administered prior to ERCP with stent placement when incomplete drainage is anticipated. Because episodes of acute cholecystitis after SEMS have been reported to be 1.9%–12%, prophylactic antibiotics may have a role, but this requires further studying.24,25 Endoscopy societies have specific recommendations for placement of PS or SEMS with regard to patient’s predicted survival. The European Society for Gastrointestinal Endoscopy and the Asia-Pacific Working Group on hepatobiliary cancers recommend SEMS if predicted survival is more than 3 months.22,26
The use of PS in biliary strictures was studied retrospectively in 70 patients (31 with PCC). Use of two stents for complete drainage of complex hilar strictures was associated with longer survival compared to drainage with only one stent (176 vs 119 days). However, PS was associated with 27% incidence of occlusion due to sludge and subsequent cholangitis and 8% dislocation rate.27 The 2013 Asia Pacific consensus for hilar cholangiocarcinoma (CC) recommends adequate drainage of ≥ 50% of the liver. This may be achieved by unilateral or bilateral stenting.28 For Bismouth-Corlette (BC) lesions III and IV, it has been shown that percutaneously placed SEMS are associated with a higher technical success rate than endoscopic SEMS (93% vs 77%).29 Mukai and colleagues demonstrated that SEMS have significantly better patency rate than PS at 6-month follow-up (80% vs 20%) and remained patent at patient’s death in 60% of the cases, compared to 30% with PS. The main reason for occlusion was sludge and tumor ingrowth for PS and SEMS, respectively.30
Many authors have argued in favor of bilateral stenting. A study of 46 patients demonstrated significantly longer patency in bilateral compared to unilateral stenting (488 vs 210 days, 23% vs 59% occlusion rates, respectively).31 Another study on PS and SEMS showed that the patency of bilateral SEMS and PS was better than unilateral approach (29 vs 24 weeks for SEMS and 18 vs 17 weeks for PS).32 That study also showed significantly more occlusion rates of PS (52%) compared to SEMS (24%), longer patency of SEMS, and more incidence of cholangitis with PS use.32 Bilateral stenting can be achieved by stent-instent (SIS) or side-by-side (SBS) techniques. The SIS and SBS approaches have overall functional success of almost 100%, technical success of 80%–100% and 73%–100%, revision rates 6%–100% and 3%–45%; and patency on average 140–217 days and 130–169 days, respectively.33 Very high technical and functional success rates and easy re-intervention were also demonstrated in a study of large cell SIS approach in a single session. The occlusion rate was 42%, mainly due to sludge and tumor ingrowth.34 Simultaneous deployment of SBS or SIS with the 6 Fr delivery system showed similar success rates, occlusion rate of 25% and re-intervention rate of 50% in about 3 months of follow-up. It is suggested that SBS or SIS should be chosen on a case-by-case basis, probably favoring SIS placement in primary sclerosing cholangitis (PSC) due to narrower ducts.35 The concept of placing bilateral stents for complete drainage of the obstructed hilum has been challenged in many studies. In the prospective study of 157 patients showed that unilateral stent placement was more technically feasible than bilateral (87% vs 77%) and was associated with less cholangitis (9% vs 17%), probably due to less contrast injection and wire manipulation of the unstented duct.36 A recent study did not find benefit in using bilateral SEMS or PS in perihilar lesions. The authors found 73% and 50% incidences of occlusion in bilateral PS and bilateral SEMS respectively.30 Another study showed higher incidence of liver abscesses with bilateral compared to unilateral stenting (18% vs 1.5%).37 A recent meta-analysis did not find benefit of bilateral stenting compared to unilateral regarding occlusion and stent failures.38 Nevertheless, in cases of cholangitis of the undrained lobe, persistent jaundice after unilateral stent placement, or when brachytherapy is anticipated, bilateral stenting should be implemented.33,39 De Palma et al40 conducted a prospective study in 61 patients with BC type II, III, and IV lesions using uncovered SEMS and showed that successful drainage was achieved in 97% of the cases. The median stent patency was 169 days, and most occlusion occurred after one month following the stent placement (23%).
Many authors have suggested that use of covered stents for hilar strictures may be associated with a high likelihood of cholangitis. The retrospective study of 36 patients with PCC showed that the use of uncovered SEMS was associated with median patency of 169 days and occlusion rates of 9% within a month and 23% after one month of stent placement.41
In order to overcome some of the technical difficulties of SBS SEMS for bilateral drainage, a Y-configuration stent with wide mesh design was developed. In a pilot study in 10 patients the Y-stent demonstrated median patency of 217 days and occlusion rate of 25% without any early adverse events.42 Multiple studies have shown excellent technical and clinical success rates with the use of Y-stents ranging between 87% and 100%. Stent occlusion rates are usually between 25% and 39%.43–46 Cholecystitis (28% late and 9% early) and cholangitis (16% late and 6% early) are reported complications of Y-stents.47 Another stent without wide mesh in the middle was evaluated in 35 patients and demonstrated 94% placement success rate, mean patency of 150 days and very low rate of cholangitis (6%). However, recent comparison of stents with large and small cells did not demonstrate any differences in tumor ingrowth or overgrowth, stent occlusion, or success in drainage.47,48 With the SIS technique, multiple stents to drain three or 4 ducts simultaneously can be placed. This approach has been associated with better patency and survival compared to one or two-branch drainage to control disease. Occlusion rates range from 33% to 68%.49–51
The role of stents for DCC is more clearly established. SEMS have larger diameter than PS and subsequently have longer patency. The 30 Fr covered SEMS were compared with 10 Fr PS in a prospective study of 100 patients (9 with CC). The patency was longer in the covered SEMS compared to the PS (3.8 and 1.8 months respectively). The reason of the short patency in that study was attributed to short survival (4.5 months). Sludge was the primary reason for PS occlusion.52 Covered and uncovered SEMS have been compared in multiple studies. Overall there is significantly higher rate of stent occlusion with uncovered stents mainly due to tumor ingrowth, ranging between 21% and 38% compared to 14%–19% with covered stents where occlusion occurs usually due to tumor overgrowth. The rate of stent migration is higher with covered stents, between 3% and 36%, compared to 0% 2% with uncovered stents. Pancreatitis is more common with covered stents (6% vs 1%) and there seems to be no association of stent type and incidence of cholecystitis.53–57 The recent multicenter study of fully covered SEMS with flared ends showed excellent success rates, median patency of 328 days and very low late adverse events, including 4 cases of SEMS migration. Removability of fully covered stents is acceptable. With an attention to place the stent below the cystic duct, the incidence of cholecystitis is very low (0.75%).58 SEMS are not associated with leak either at the anastomosis or the surgical margins, and do not affect 30-day mortality or length of stay.59,60 The recent meta-analysis of 2,000 patients with distal and proximal strictures demonstrated that SEMS have lower early and late occlusion rates for both hilar and distal lesions compared to PS. There was 13% therapeutic failure with PS compared to 7% with SEMS and less need for re-interventions with the latter. Cholangitis is more common with PS than with SEMS (21% vs 8%). Stent migration is non-existent with uncovered SEMS, 12% with partially covered and 6.5% with PS.38 Placement of stents under EUS guidance with an anterograde approach (rendezvous) has been shown to be technically (90% 100%) and clinically successful (80% 100%). Unsuccessful cases may be due to inability to traverse the biliary stricture, presence of choledochocele, failed guidewire passage, or loss of scope position. Adverse events have been documented in about 15% of the cases. The EUS-guided anterograde or choledochoduodenal approach was equally successful with the ERCP approach (94% vs 93%) with similar incidence of adverse events, except for pancreatitis, which was higher with ERCP (5% vs 0%).61–63
Surgery for intrahepatic cholangiocarcinoma (ICC) is usually not indicated in tumors that invade the vasculature, multiple tumors, bilobar disease, and metastases. Regional lymph node (LN) involvement is not a contraindication for surgery.1 Child-Pugh class B and C, Model for End Stage Liver Disease score > 9, or portal hypertension are relative contraindications.64 There is not enough data to suggest routine staging laparoscopy prior to surgery.65 Preoperative PV embolization may be performed if FLR is anticipated to be less than 30%.66 A recent multinational study calculated the median survival after resection of ICC at 27 months.67 The survival rate after 1, 3, and 5 years was 75%, 39%, and 25%, respectively. For American Joint Committee on Cancer 7th stage I and II lesions, the overall survival rate was 33% vs 16% for stages III and IV. Patients with and without LN metastasis had 5-year survival rates of 11% and 35% respectively. Vascular invasion decreased 5-year survival rate to from 28% to 16%. Tumor size ≥ 5 cm in size was associated with a survival rate of 22% at 5 years compared to that of 33% for tumors < 5 cm. Multiple lesions were associated with a poor 5-year survival rate of 17% compared to 29% in those patients with only one lesion. Interestingly, the presence of cirrhosis did not influence survival. The study calculated the conditional probability for survival in patients with ICC. Among other interesting data, it was shown that patients who survived 3 years after index operation had 53% and 39% chance to be alive at 5 and 7 years respectively.67 Extended liver resection is frequently attempted in order to resect the whole tumor with R0 margins. Extended resections have been associated with increased postoperative mortality and bile leaks compared to more conservative resections.68,69 Readmission rate after surgery is 8%–10% and is higher in patients with preoperative jaundice and major postsurgical adverse events such as bilomas, infections and pleural effusions.70 Recurrence of ICC, even after R0 resections, is high with the main site of recurrence being the liver.71,72
LN dissection and extended lymphadenectomy for ICC are a topic of debate. For staging and exclusion of positive regional LN, dissection may be indicated in patients with mass forming (MF) type and mixed MF + periductal infiltrating (PI) type of ICC. LN dissection confers no survival benefit if there are no intrahepatic metastases or clinically negative LN.73,74 LN metastases has been shown to be more common in MF, PI, or mixed types and less in the intraductal growth type of ICC. Well differentiated tumors and carbohydrate antigen 19-9 less than 135 U/mL seem to be associated with less frequent LN involvement.75 Patients with solitary, small tumor (< 5 cm), and peripherally located tumors may not require LN dissection.76
Surgery is the only way to achieve cure in PCC. Criteria for unresectability have been proposed to be 1) distant metastases, 2) N2 stage, 3) bilateral proximal ductal involvement deep above the second order ducts, 4) involvement of the common hepatic artery (HA) or PV main trunk, 5) unilateral involvement of second order ducts with contralateral vascular involvement, 6) lobar atrophy with contralateral involvement of second order ducts, and 7) lobar atrophy with involvement of contralateral PV or HA.77 The surgical approach for PCC has changed significantly over the last decades. Currently the followed surgical approach is major hepatectomy, caudate segmentectomy, and extrahepatic bile duct resection.78,79 However, more radical surgeries with pancreatoduodenectomy (PD) have been proposed.80 Data from various studies indicate that the median 5-year survival rate after resection of PCC is about 11% 44% in studies done until 2010 and 14% 66% thereafter, with surgical mortality 0%–23% and 0%–14%, respectively. A recent meta-analysis found mortality after hepatectomy for PCC of 3.7%.81,82 The 5-year survival, although longer, has not been significantly different in R0 and R1 resections.83,84 Achievement of R0 and R1 surgical margins in non-metastatic disease has been associated with median survival of 40–74 months and 9–24 months respectively in some studies. For R2 resection, the 3year survival drops significantly (7%) and it is extremely poor for unresectable tumors (9% at 12 months).13,85–88 The goal of surgery is to resect as much liver as possible in order to achieve R0 margins and maintain more than 40% remnant liver after hepatectomy. Therefore, preoperative PV embolization is the standard of care before hepatectomy to minimize postoperative liver failure.82 It has been shown that macroscopic tumor invasion of the PV impacts survival.89 However, PV resection and reconstruction (PVR) remain debatable. de Jong et al90 reported that PVR did not improve the 5-year survival (28% vs 33% in those without PVR), but was associated with an increased 30- and 90-day mortality. When the tumor invades the PV or the HA, vascular resection with reconstruction can be attempted. The recent meta-analysis of 669 patients who underwent vascular resection for PCC demonstrated higher mortality in those who had HA resection and reconstruction, with adverse events, such as thrombosis, hemorrhage, pseudoaneurysms, and anastomotic leaks.91 PVR is associated with adverse events such as late anastomotic strictures, thrombosis, and variceal bleeding.92 The right trisectionectomy combined with PVR
LN sampling is indicated for PCC staging, but its utility in curative resection is still debatable.94 The incidence of LN involvement in PCC ranges between 24% to more than 50%. Presence of LN metastasis and the number of malignant LN are associated with survival.95,96 A recent meta-analysis reported higher rate of LN metastases in patients who underwent PVR compared to those who did not, and significantly poorer overall survival.97
Due to its anatomic location, DCC may involve the head of the pancreas or may be confused for pancreatic cancer.98 The surgical procedure for DCC is PD. Tumors in the middle of the common bile duct may be managed with duct resection and Roux-en-Y hepaticojejunostomy.99 Surgery is generally not recommended in the presence of distant metastases, LN involvement beyond the PV, HA and peripancreatic and celiac axis, more than 180 degrees involvement of the HA and superior mesenteric artery, and if more than 2 cm of resection is anticipated in cases of portal and superior mesenteric vein involvement.100 The role of staging laparoscopy for DCC is along the lines of ICC and PCC, and has not been clearly established.99 Postoperative survival rate for DCC has been reported between 18%–54%.101 Twenty-three percent 5-year survival rate and median survival of 18 months were reported in a series of 563 patients.102 Surgical mortality in large published series has been less than 3%. Adverse outcomes after PD are pancreatic leaks, delayed gastric emptying, abscesses, bile leaks, sepsis, surgical site infections, and cardiopulmonary adverse events.102,103 For very aggressive tumors involving the common bile duct from the ampulla to the hepatic hilum, hepatectomy combined with PD has been suggested to be the recommended approach. A mortality of 2.4% was reported in a recent series, but the procedure was accompanied by a high morbidity.80 Palliative biliary drainage for DCC and PCC is in favor of the endoscopic approach, so surgical biliary bypass may be indicated only for patients who fail stenting.104
A recent multicenter study from Japan assessed the prognostic impact of LN metastases in DCC.105 It was demonstrated that in DCC, LNs were involved in 42% of the cases and the overall survival was 53%, 41%, and 28% at 3, 5, and 10 years, respectively. The mean survival without LN involvement was 5.7 years vs 1.9 years for patients with positive LN. Four or more positive LNs were associated with low survival (median, 1.3 years vs 2.2 years with less than 4 LN).105 Factors that affect postoperative survival are adjuvant chemotherapy, higher degree of differentiation, tumor size < 2 cm, R0 margins, negative LN, and earlier stage of disease.101,102
At this time, liver transplantation (LT) for ICC should be considered within research trials along with neoadjuvant and/or adjuvant therapy.65 Survival benefit of LT for ICC does not reach the rate achieved for HCC and thus it is perceived with skepticism among physicians.106 The 5-year survival rates for ICC managed with LT have been reported to be suboptimal.106 A small study showed 5-year survival of 33% and a larger one demonstrated 29% survival rate; however, the latter included large tumors (≤ 8 cm).107 In patients with cirrhosis and ICC, it has been shown that LT is associated with poorer outcomes compared to LT for HCC (5-year survival rate 51%). However, for uninodular and smaller than 2 cm tumors (including mixed HCC–ICC), LT delivered much better survival results (62%–73%).108,109
LT is indicated in patients with unresectable PCC after administration of neoadjuvant chemoradiotherapy. Staging laparoscopy with LN sampling is performed to exclude LN malignant infiltration, which is an absolute contraindication for LT. Recent studies have demonstrated good 5-year survival rates of around 65%–82%.110–113 Surgical or percutaneous biopsy of the tumor is a contraindication for LT due to the risk of tumor seeding.114 In patients with PSC, who are rarely optimal candidates for R0 resection due to multifocal nature of the tumor, LT preceded by neoadjuvant protocol is the treatment of choice. The decision of resectionvs LT is often challenging. Attempted surgical resection for PCC deems a patient non-transplantable if LNs are positive intraoperatively or R0 margins are not achieved postoperatively. Also, neoadjuvant radiation protocol for LT induces liver damage to such extent that probably precludes surgical resection.115 In patients with
The role of neo-adjuvant chemotherapy for CC is not established and has an unclear benefit. Although solid evidence is lacking, its use aims to downsize the tumor and increase the possibility of R0 resection, as shown with neo-adjuvant gemcitabine.117 Pathologic response has been rarely reported in the literature with combinations of gemcitabine, cisplatin, doxorubicin, or 5-fluorouracil (5-FU).118–120 Neo-adjuvant chemoradiotherapy for advanced, resectable PCC and DCC has yielded good results with improvement in 5-year survival compared to adjuvant therapy alone.121
Adjuvant therapy for ICC is a matter of debate. It may be of benefit in cases of lymphovascular and perineural invasion, positive surgical margins, LN involvement, and tumor size above 5 cm. According to the National Comprehensive Cancer Network® (NCCN®) (hepatobiliary cancers version 2.2016), R0 resections may not benefit from adjuvant therapy, so observation is an option and participation in a clinical trial is encouranged (Table 2).122
Adjuvant chemotherapy and radiotherapy for ICC have been associated with statistically improved survival in patients with postoperative positive LN and surgical margins, but not in those with negative LN and surgical margins.123 A systematic review found the combination of adjuvant chemotherapy and radiotherapy was beneficial in the minority of the patients.124 A meta-analysis demonstrated significant improvement in survival with the adjuvant combination of chemotherapy and radiotherapy than adjuvant radiotherapy or surgery alone.125 Recent data from the National Cancer Database demonstrated that for patients with R1/R2 resection, LN involvement, and T3–T4 stage of disease, adjuvant chemotherapy improved the overall survival compared to surgery alone (20 vs 11 months for LN positive patients, 21 vs 16 months for advanced stage, and 19.5 vs 12 months for R1/R2 resection, respectively).126 The meta-analysis of the prospective ABC-02 and BT-22 trials comparing gemcitabine plus cisplatin combination to gemcitabine monotherapy for biliary malignancies, concluded that the combination significantly improved the performance status and overall survival (11.6 vs 8 months).127 A study from MD Anderson Cancer Center on patients with extra-hepatic CC (ECC) demonstrated equal survival in R0N0 patients treated with surgery alone and R1N1 patients treated with surgery and chemoradiation, suggesting a benefit of adjuvant chemoradiation in these patients.128 Adjuvant radiotherapy for ECC was found to be well tolerated and to have a significant impact on survival in a recent meta-anaysis.129 A phase II trial in postoperative R0 and R1 patients with CC and gallbladder cancer showed that capecitabine/gemcitabine followed by capecitabine and radiotherapy was associated with an overall 2-year survival rate of 65% and disease-free survival rate of 51%.130
The NCCN recommends biliary drainage in patients with jaundice if chemotherapy is planned (Table 3).122 A study of almost 400 patients with either unresectable or metastatic cholangiocarcinoma treated with cisplatin and gemcitabine vs gemictabine alone demonstrated a significant improvement in overall survival with the combination therapy (12 and 8 months, respectively) and better progression-free survival (8 vs 5 months).131 The recent pooled analysis of 161 trials demonstrated that gemcitabine-platinum therapy was associated with overall survival of 9.5 months and time-to-tumor growth of 5 months. Addition of fluoropyrimidine or gemcitabine based chemotherapy plus targeted therapy mainly against epidermal growth factor receptor yielded even better outcomes compared to gemcitabine platinum alone.132
External-beam radiotherapy (EBRT) and brachytherapy are used in patients with unresectable CC and may have a role in relieving jaundice and pain.65 EBRT in unresectable ICC has been shown to benefit the overall survival. In a retrospective study, survival with radiotherapy was 7 months compared to 3 months without any treatment, whereas surgery plus radiotherapy conferred 11-month survival.133 EBRT may downsize the tumor, but is associated with radiation injury to the adjacent organs.134 Two studies of radiotherapy in unresectable ICC have demonstrated a good local disease control and improved survival (13 months), however associated with significant toxicity (liver failure, biliary stenosis).135,136
Stereotactic body radiation (SBRT) delivers higher doses of radiation in attempt to control tumor progression. Small studies have shown good local disease control and potential survival benefit of SBRT alone or combined with EBRT for unresectable and recurrent CC.136,137 To minimize the radiation-induced toxicity, treatment with proton beams has been suggested. A recent study demonstrated that after percutaneous implantation of fiducial markers for localization, proton beam therapy was associated with an increased median survival compared to palliative care (27.5 vs 10 months), making it a potentially promising modality in the management of locally advanced ICC.138
Brachytherapy is a form of radiotherapy that may be used to treat postoperative local CC recurrence or as palliative treatment in unresectable CC. Due to its local effect, it is less toxic to the adjacent organs compared to other forms of radiotherapy. It has been shown to improve the biliary stent patency and maybe the overall survival.139 The combination of brachytherapy and EBRT has been shown to be better than brachytherapy alone.140 Brachytherapy with iridium71 plus SEMS placement was shown to improve survival from PCC by 90 days compared to SEMS alone.141 A retrospective study demonstrated that brachytherapy, with or without EBRT, was associated with significantly improved survival (11 vs 4 months) compared to patients who received no radiation for ICC and ECC (distal and hilar).142 Iodine125 seeds along with biliary stents have been used for brachytherapy in patients with PCC, without causing obstruction of the ducts.143 The recent study from Mayo Clinic showed that brachytherapy may be administered safely via endoscopically placed nasobiliary tubes in patients with unresectable CC.144
Hepatic artery infusion (HAI) delivers the drug to the tumor after implantation of a pump or port. Combination of weekly pegylated interferon α-2b and intra-arterial infusion of 5-FU has been shown to confer 1-year survival rate of 54% and median survival of almost 15 months in a recent study.145 The recent phase I/II study from Japan showed that gemcitabine infusion was associated with tumor response of 8%, with neutropenia being the major side effect of this treatment. HAI has been reported with 5-FU-based regimens and has shown effectiveness, but large trials are lacking.146
Transarterial chemoembolization (TACE) is another approach for unresectable ICC that targets the tumor with chemotherapeutic drugs followed by embolization. It is usually contraindicated in the setting of PV thrombosis.147 A large study comparing triple combination of gemcitabine plus mitomycin C plus cisplatin to either dual-drug or single-drug approach showed partial tumor response of 9%, median survival of 13 months, but no statistical difference among all drug combinations.148 A similar median survival was demonstrated with mitomycin C (11 months), comparable to the survival of patients with positive LN or positive surgical margins.149 Treatment of unresectable ICC with doxorubicin-eluting beads has shown 100% tumor response rate and median survival of 13 months.150 TACE with irinotecan eluting beads has shown a better progression-free survival and overall survival compared to conventional TACE with mitomycin C, and comparable results to gemcitabine/oxaliplatin chemotherapy (survival 12, 6, and 11 months, respectively), with no observed hematological side effects.151
The retrospective analysis of 20 studies compared the arterial-based therapies for ICC.152 It demonstrated that HAI was associated with the longest median overall survival compared to Yttrium-90 chemoembolization, TACE, and drug-eluting bead TACE (23, 14, 12, and 12 months, respectively). Patients treated with HAI responded better compared to the other modalities, however with higher rate of adverse events.152 Radioembolization with Yttrium-90 microspheres is a form of transarterial brachytherapy for ICC. The recent pooled analysis of current studies demonstrated median survival of 15.5 months, partial tumor response in 28%, and no progression in 54% of the cases at 3 months. Side effects usually include abdominal pain, fever, and nausea, but ascites and ulcers may occur as well.153 Yttrium-90 radioembolization with systemic chemotherapy shows a promise in downstaging unresectable ICC for attempted R0 resection. The largest study to-date demonstrated successful R0 resections in all cases, mostly after extended hepatectomy. In carefully selected patients with good performance status and liver function, this might be a reasonable approach.154
Radiofrequency ablation (RFA) is a minimally invasive modality for inoperable CC. The recent meta-analysis of published data on RFA in ICC demonstrated 1-, 3-, and 5-year survival rates of 82%, 47%, and 24%, respectively. Ablated tumor size ranged between 0.7 cm and 10 cm. Most of these tumors were treated with ultrasound-guided RFA, which might be the reason for residual disease in cases of large-size primary tumor. In cases where stereotactic RFA was used, even tumors as large as 10 cm were effectively ablated without residual disease. Potential side effects of RFA include liver abscesses, pleural effusions, and bleeding. Subcapsular location and proximity of the tumor to major vessels influence decision for RFA.155 RFA may be delivered percutaneously or endoscopically through a duodenoscope. It may be used to reduce tumor burden in ECC and prolong patency of biliary SEMS. A recent study showed that percutaneous RFA and SEMS for DCC was associated with median SEMS patency of 149 days and median survival of 181 days respectively.156 A retrospective European study demonstrated median stent patency of 170 days and overall survival of 11 months after RFA treatment in patients with PCC. Hepatic infarct is a rare complication of RFA.157 Another study from the United States showed that endoscopically-administered RFA improved stricture diameter, though more pronounced for pancreatic cancer-related strictures.158 RFA every 3 months had a complete technical success and stent patency (with PS or SEMS) of 66% at 1 month after RFA.158 The role of RFA in clearing occluded SEMS needs further studying.159 ERCP-guided RFA was compared with photodynamic therapy (PDT) in patients with unresectable CC and was found to have an equal overall survival, stent migration rate, incidence of abscesses, and need for PTBD, but more episodes of stent occlusion and cholangitis than PDT.160
PDT is a technique that uses laser irradiation after administration of a photosensitizer. The patient needs to be aware that exposure to sunlight may cause skin blisters, pruritus, and erythema, and should avoid exposure 4–6 weeks after PDT.161 PDT may be applied endoscopically or percutateously in patients with unresectable PCC and DCC, patients with anticipated R1 or R2 resection, or poor surgical candidates. Due to limited tissue penetration, deep lesions may not be appropriate for PDT. Nondraining liver segments should not be treated with PDT.162,163 PDT with cholangioscopy is an attractive approach to deliver therapy precisely at the tumor site.164,165 Biliary stenting followed by PDT has been shown to prolong survival compared to stenting alone (493 and 98 days, respectively).166 The study from the United States demonstrated a similar survival benefit (16 vs 7 months).167 Stent patency was also improved with use of PDT.168
This is definitely an exciting time for CC. Currently, the diagnosis and management are challenging and the outcomes are frequently disappointing for the patient and physician. Early diagnosis and surgery may prove curative, but this is rarely the case. Advances in imaging will certainly improve preoperative staging. EUS-FNA is the best modality to assess the status of LN. A combination of brush cytology, biopsy and fluorescence
© The Society of Gastrointestinal Intervention. Powered by INFOrang Co., Ltd.