Comparison of prognostic factors and their differences in intrahepatic, hilar, and distal cholangiocarcinoma: A systematic review and meta-analysis

Abstract

BACKGROUND

Cholangiocarcinoma (CCA) comprises heterogeneous malignancies arising at different anatomical locations: Intrahepatic cholangiocarcinoma (IHCC), perihilar cholangiocarcinoma (PHCC), and distal cholangiocarcinoma (DCC). These subtypes exhibit distinct clinical behaviors, treatment approaches, and outcomes. Despite advances in surgical and adjuvant therapies, the prognostic implications of tumor location remain unclear and inconsistently reported. Understanding these variations is essential for personalized management and staging refinement. We hypothesized that the anatomical subtype of CCA significantly influences prognostic outcomes and pathological features.

AIM

To compare prognostic outcomes and clinicopathological characteristics among IHCC, PHCC, and DCC based on current evidence.

METHODS

A systematic review and meta-analysis were conducted in accordance with PRISMA guidelines. PubMed, EMBASE, and the Cochrane Library were searched, yielding 11 eligible retrospective comparative studies involving 14484 patients (IHCC: 6260; PHCC: 6895; DCC: 1329). Outcomes assessed included overall survival (OS), lymph node metastasis, neural invasion, and vascular invasion. Statistical analyses were performed using RevMan 5.3 and Stata 13.0.

RESULTS

DCC demonstrated the most favorable prognosis among all subtypes. Despite the highest lymph node metastasis rate (DCC: 56.9%), it was associated with better OS than PHCC and IHCC. Vascular invasion was more prevalent in IHCC (OR = 1.66, 95%CI: 1.22-2.28, P = 0.001). OS comparisons showed no significant difference between PHCC and IHCC (HR = 1.02, P = 0.88), while DCC showed consistent trends toward better survival against both.

CONCLUSION

Anatomical subtype is a significant prognostic factor in CCA. DCC patients experience superior outcomes despite aggressive lymphatic spread, suggesting better resectability and surgical outcomes. These insights underscore the need for subtype-specific management strategies and future prospective validation.

Key Words: Cholangiocarcinoma; Intrahepatic; Perihilar; Distal; Prognosis; Lymph node metastasis; Survival; Meta-analysis

Core Tip: This meta-analysis offers a comprehensive comparison of intrahepatic cholangiocarcinoma (IHCC), perihilar cholangiocarcinoma (PHCC), and distal cholangiocarcinoma (DCC), revealing significant prognostic differences. DCC shows the most favorable survival outcomes despite a higher rate of lymphatic invasion. PHCC has a better prognosis than IHCC, although both share similar tumor biology. These findings underscore the heterogeneity of cholangiocarcinoma and support the need for tailored treatment strategies based on tumor location.

INTRODUCTION

Cholangiocarcinoma (CCA) is an aggressive biliary epithelial cancer that can develop anywhere along the biliary tract, from the intrahepatic biliary canaliculi to the duodenal ampulla, where the common bile duct enters the duodenum. CCA is categorized into two types based on its anatomical origin: Intrahepatic CCA (iCCA) and extrahepatic CCA, which is further differentiated into perihilar CCA and distal CCA (dCCA). Over 95% of CCAs are adenocarcinomas[1]. CCA is a heterogeneous, uncommon malignancy with a prevalence of around 1.26 per 100000 persons[2]. iCCA develops from intrahepatic small or large bile ducts, including the second-order bile ducts. 15%-20% of all bile duct cancers are iCCAs, which have an extremely bad prognosis[3]. Hilar CCA is a biliary system cancer that affects the proximal extrahepatic bile ducts. It accounts for 50%-70% of all bile tract neoplasms while being an uncommon illness[4]. dCCA occurs beyond the junction where the cystic duct enters the common hepatic duct to create the common bile duct and is proximal to the Vater's ampulla. Approximately 30% of CCA manifests as dCCA[5-7]. Prognostic and clinical characteristics must be compared among intrahepatic, hilar, and dCCA subtypes due to their unique anatomical placements, biological behaviors, and treatment responses. These comparisons allow for the discovery of subtype-specific prognostic markers, which facilitates personalized therapy regimens and improves patient outcomes. For example, a research examining prognostic markers following surgical resection for several CCA subtypes found disparities in survival rates and recurrence patterns, emphasizing the need for distinct therapy methods[8]. Furthermore, recognizing these differences assists in the creation of precise staging systems and improves the accuracy of survival forecasts across various CCAs[9,10]. So that’s why we are performing this meta analysis to evaluate the differences in prognostic and clinical factors among intrahepatic (ICC), hilar, and distal cholangiocarcinoma (DCC).

MATERIALS AND METHODS

Search strategy

Our meta-analysis was performed based on The PRISMA Statement for Reporting Systematic Reviews and Meta-Analyses of Studies That Evaluate Health Care Interventions: Explanation and Elaboration. We thoroughly searched databases (PubMed, EMBASE and the Cochrane library) till January 1, 2025[11]. Eligible studies are restricted to comparative studies between different kind of CCA according to different locations. The following keyword were used for literature searching: (cholangiocarcinoma) OR (bile duct cancer) AND (hilar) AND (distal) OR (intrahepatic) AND (distal) OR (intrahepatic) AND (hilar).

Inclusion criteria

(1) English literature; (2) Comparative studies (two-arm or triple-arm) regarding different types of CCA based on its tumor locations (intrahepatic, hilar and distal); (3) Studies reported the pathological features or the long-term prognosis; and (4) Studies were included if they were published in peer-reviewed journals.

Exclusion criteria

(1) Studies with inadequate original date for further analysis; (2) Studies came from the same institute or shared a completely same patient source; (3) Abstracts, letters, meeting conference or reviews; and (4) Single-arm studies.

Data extraction

Data extraction was conducted systematically using a standardized form to ensure consistency and accuracy. Two independent reviewers extracted relevant data from each included study, resolving discrepancies through discussion or consultation with corresponding author when necessary. Data required in our analysis were either estimated or calculated or directly extracted from the original studies. The basic characteristics of all available studies were summarized in Table 1, including author, country, study period, study design, the number of patients in each group, patient source, tumor classification criteria and follow up (months). A total of five measured outcomes were finally identified. Overall survival (OS) was regarded as the primary measured outcome including lymph node metastasis, neural invasion and vascular invasion. The pooled results of all available studies in measured outcomes were summarized in Table 2.

Table 1 Baseline characteristics of all studies included.

Ref.	Country	Study period	Study design	No. patients based on tumor location			Patient source	Tumor classification criteria	Follow up (month)	Quality score (NOS)
				IHCC	PHCC	DCC
DeOliveira et al[15]	America	1973-2004	Triple-arm	44	281	239	Johns Hopkins Hospital	NA	Median follow up: 29	7
Murakami et al[8]	Japan	1990-2009	Triple-arm	21	50	56	Hiroshima University Hospital	UICC, 7th edition	Median follow up: 78 (2-254)	7
Yusoff et al[16]	Malaysia	1997-2007	Triple-arm	12	38	19	A tertiary referral center in Malaysia	TNM staging, 2002	NA	7
Ercolani et al[17]	Italy	1985-2011	Triple-arm	136	150	53	Two Italian tertiary referral University Centers	AJCC, 7th edition	Median follow up: 30.2 ± 38	8
Waseem and Tushar[18]	America	1992-2010	Triple-arm	90	104	48	Mayo Clinic	UICC, 7th edition	Median follow up: 11.6 (3.1-24.3)	9
Guglielmi et al[19]	Italy	1990-2007	Two-arm	33	62	None	University of Verona Medical School	AJCC, 7th edition	Every 6 months	8
Hang et al[20]	China	1973-2015	Triple-arm	5462	5626	622	SEER database + TCGA database	NA	NA	7
Nakanishi et al[21]	Japan	1998-2018	Two-arm	None	236	130	Hokkaido University Hospital	AJCC, 8th edition	Every 3–6 months	8
Sallinen et al[22]	Finland	2000-2015	Two-arm	None	36	47	A tertiary hospital in Finland	AJCC, 7th edition	NA	7
Istanbouli et al[23]	America	1995-2016	Triple-arm	64	75	31	Mayo Clinic	AJCC, 7th edition	NA	7
Sarkhampee et al[24]	Thailand	2013-2018	Triple-arm	398	237	85	Sunpasitthiprasong Hospital	AJCC, 7th edition	3 to 6 months during the first 2 years. Then every 6 to 12 months thereafter	8

IHCC: Intrahepatic cholangiocarcinoma; PHCC: Peri-hilar cholangiocarcinoma; DCC: Distal cholangiocarcinoma; UICC: International Union Against Cancer; AJCC: American Joint Committee on Cancer; NOS: Newcastle Ottawa Scale.

Table 2 Pooled results of all available studies in measured outcomes.

Measured outcomes	No. studies	No. patients		Model (fixed/random)	OR/HR	95%CI	P (overall test)	Heterogeneity		Begg test	Egger test
								I2	P value	Pr >|z1	P >|t|1
Overall survival	9	IHCC: 6296	PHCC: 6716	Random	HR = 1.02	0.83-1.24	P = 0.88	58%	0.01	1	0.137
Node metastasis	6	IHCC: 651	PHCC: 817	Fixed	OR = 0.69	0.55-0.88	P = 0.003	0%	0.85	0.469	0.492
Neural invasion	3	IHCC: 435	PHCC: 382	Random	OR = 0.71	0.11-4.64	P = 0.72	96%	< 0.00001	1	0.671
Vascular invasion	3	IHCC: 453	PHCC: 384	Random	OR = 0.82	0.28-2.38	P = 0.72	90%	< 0.0001	1	0.657
Overall survival	8	IHCC: 6227	DCC: 1153	Random	HR = 1.07	0.18-1.36	P = 0.59	56%	0.03	0.533	0.355
Node metastasis	5	IHCC: 522	DCC: 445	Fixed	OR = 0.33	0.23-0.45	P = 0.00001	70%	0.009	0.816	0.631
Neural invasion	-	-	-	-	-	-	-	-	-	-	-
Vascular invasion	3	IHCC: 661	DCC: 250	Fixed	OR = 1.66	1.22-2.28	P = 0.001	91%	< 0.0001	0.333	0.263
Overall survival	10	PHCC: 6833	DCC: 1330	Random	HR = 1.06	0.92-1.22	P = 0.42	41%	0.09	1	0.352
Node metastasis	7	PHCC: 989	DCC: 622	Random	OR = 0.48	0.29-0.81	P = 0.005	80%	< 0.0001	0.381	0.952
Neural invasion	3	PHCC: 573	DCC: 250	Random	OR = 2.35	0.76-7.28	P = 0.14	90%	< 0.0001	0.333	0.065
Vascular invasion	-	-	-	-	-	-	-	-	-	-	-

¹P value for statistical tests (e.g., Begg and Egger tests).

IHCC: Intrahepatic cholangiocarcinoma; PHCC: Peri-hilar cholangiocarcinoma; DCC: Distal cholangiocarcinoma.

The extracted data were then compiled into a comprehensive database for analysis, allowing for robust comparisons of prognostic factors across intrahepatic, hilar, and dCCA subtypes.

Quality assessment

The quality of all identified studies were evaluated by author 1 and author 2 according to the Newcastle-Ottawa Scale score[12]. The quality scores > 6 indicated that the quality of the cohort was relatively high. The results were presented in Table 1.

Statistical analysis

RevMan5.3 software and Stata 13.0 software were used for statistical analysis. The OR was applied in the analysis of dichotomous variables. Pooled estimates for OS rates and HR were calculated using a random-effects model to account for variability between studies. Heterogeneity was assessed using the I² statistic, with values greater than 50% indicating significant heterogeneity[13]. Subgroup analyses were performed based on tumor locations to explore differences in outcomes. Sensitivity analyses were conducted to evaluate the robustness of the findings. A P value of less than 0.05 was considered statistically significant.

Begg’s and Egger’s tests were used for evaluating the potential publication bias within comparisons. A P value or corrected P value lower than 0.05 indicated the existence of significant publication bias[14]. Begg’s funnel plot and Egger’s liner regression were also applied in furtherly valiating the publication bias with the comparison of OS.

RESULTS

Study identification and selection

Initially, based on our search strategy, 4941 relevant articles were initially identified. According to the inclusion criteria, 11 comparative studies were finally included. The specific process is presented in Figure 1.

Figure 1  The specific process of literature researching and selection.

Study characteristics

A total of 11 retrospective cohort studies[8,15-24]. Eight studies[8,15-18,20,23,24] were triple-arm studies and the other three studies[19,21,22] were two-arm studies. There were 6260 patients with iCCA, 6895 patients with peri-hilar CCA and 1329 patients with dCCA. All the studies were incorporated in the survival analysis.

Intrahepatic cholangiocarcinoma vs perihilar cholangiocarcinoma

OS: Nine studies[8,15-20,23,24] were incorporated into the survival analysis between intrahepatic cholangiocarcinoma (IHCC) and perihilar cholangiocarcinoma (PHCC). Meta-analysis of data from studies using a random-effects model revealed no significant difference between IHCC and PHCC (HR = 1.02, 95%CI: 0.83-1.24, P = 0.88) (χ² = 19.16, P = 0.01, I² = 58%) (Figure 2A).

Figure 2 Forest plots comparing the long term survival and the tumor biological features between intrahepatic cholangiocarcinoma and perihilar cholangiocarcinoma. A: Overall survival between intrahepatic cholangiocarcinoma vs perihilar cholangiocarcinoma; B: Node metastasis; C: Neural invasion; D: Vascular invasion. PHCC: Perihilar cholangiocarcinoma; IHCC: Intrahepatic cholangiocarcinoma.

Node metastasis: Meta-analysis of data from six studies[8,15,17,19,23,24] using a fixed-effects model revealed no significant difference between IHCC and PHCC groups (25.3% vs 30.9%, (OR = 0.69, 95%CI: 0.55-0.88, P = 0.003) (χ² = 1.98, P = 0.85, I² = 0%) (Figure 2B).

Neural invasion: Meta-analysis of data from three studies[17,19,24] using a random-effects model revealed no significant difference between IHCC and PHCC groups (49.3% vs 40.2%, OR = 0.71, 95%CI: 0.11-4.64, P = 0.72) (χ² = 54.07, P < 0.00001, I² = 96%) (Figure 2C).

Vascular invasion: Meta-analysis of data from three studies[17,19,24] using a random-effects model revealed no significant difference between IHCC and PHCC groups (60.4% vs 53.2%, OR = 0.82, 95%CI: 0.28-2.38, P = 0.72) (χ² = 20.50, P < 0.0001, I² = 90%) (Figure 2D).

IHCC vs DCC

OS: Eight studies[8,15-18,20,23,24] were incorporated into the survival analysis between IHCC and DCC. Meta-analysis of data from six studies using a random-effects model revealed no significant difference between IHCC and PHCC (HR = 1.07, 95%CI: 0.84-1.36, P = 0.59) (χ² = 15.80, P = 0.03, I² = 56%) (Figure 3A).

Figure 3 Forest plots comparing the long term survival and the tumor biological features between intrahepatic cholangiocarcinoma and distal cholangiocarcinoma. A: Overall survival between intrahepatic cholangiocarcinoma vs distal cholangiocarcinoma; B: Node metastasis; C: Vascular invasion. IHCC: Intrahepatic cholangiocarcinoma; DCC: Distal cholangiocarcinoma.

Node metastasis: Meta-analysis of data from five studies[8,15,17,23,24] using a random-effects model revealed a significantly higher incidence of node metastasis in patients with DCC (28.3% vs 56.9%, OR = 0.33, 95%CI: 0.23-0.45, P < 0.00001) (χ² = 13.55, P = 0.009, I² = 70%) (Figure 3B).

Vascular invasion: Meta-analysis of data from three studies[17,21,24] using a fixed-effects model revealed no significant difference between IHCC and DCC groups (47.5% vs 47.0%, OR = 1.66, 95%CI: 1.22-2.28, P = 0.001) (χ² = 21.43, P < 0.0001, I² = 91%) (Figure 3C).

PHCC vs DCC

OS: Ten studies[8,15-18,20-24] were incorporated into the survival analysis between PHCC and DCC. Meta-analysis of data from eight studies using a fixed-effects model revealed a significant difference between PHCC and DCC (HR = 1.06, 95%CI: 0.92-1.22, P = 0.42) (χ² = 15.19, P = 0.09, I²= 41%) (Figure 4A).

Figure 4 Forest plots comparing the long term survival and the tumor biological features between perihilar cholangiocarcinoma and distal cholangiocarcinoma. A: Overall survival between perihilar cholangiocarcinoma and distal cholangiocarcinoma; B: Node metastasis; C: Neural invasion. PHCC: Perihilar cholangiocarcinoma; DCC: Distal cholangiocarcinoma.

Node metastasis: Meta-analysis of data from seven studies[8,15,17,21-24] using a random-effects model revealed a significantly higher incidence of node metastasis in patients with DCC (33.4% vs 53.7%, OR = 0.48, 95%CI: 0.29-0.81, P = 0.005) (χ² = 30.46, P < 0.0001, I² = 80%) (Figure 4B).

Neural invasion: Meta-analysis of data from three studies[17,21,24] using a random-effects model revealed no significant difference between PHCC and DCC groups (82.1% vs 93.5%, OR = 2.35, 95%CI: 0.76-7.28, P = 0.14) (χ² = 19.88, P < 0.0001, I² = 90%) (Figure 4C).

Publication bias

We assessed publication bias using both Egger’s test and Begg’s test for each comparison. Egger’s test showed no significant small-study effects (P > 0.05 for all comparisons), and Begg’s test further confirmed the absence of publication bias. Additionally, visual inspection of funnel plots did not indicate substantial asymmetry. These findings suggest that publication bias is unlikely to have influenced our results.

DISCUSSION

According to a 2012 expert consensus that compiled data from many published series, 20% of all CCA cases were intrahepatic, 50% were peri-hilar (including the bifurcation), and 20% were distal[25]. The authors suggested that they may be categorized into three distinct entities based on their variations in frequency, clinical characteristics, and treatment[25,26]. According to the most recent National Comprehensive Cancer Network recommendations[27], the three entities' general surgical principles and the adjuvant therapy that followed were significantly different. Our meta-analysis assesses the variations among three unique CCA types in terms of long-term survival, and there is a growing tendency toward a thorough understanding of the distinguishing characteristics of these three forms of CCAs. The following are our main conclusions: (1) Patients with DCC had a far better prognosis than those with IHCC or PHCC; (2) Patients with PHCC showed a considerably better outcome than those with IHCC; and (3) Patients with IHCC and PHCC had comparable tumour pathology characteristics, however lymphatic invasion was most commonly found in DCC patients.

The first comprehensive comparative analysis, published in 2007, contained three forms of CCAs (IHCC: 44, PHCC: 281 and DCC: 239) and was based on an aged western cohort (1973-2004). All of these patients had surgical exploration, with the majority (52% of them) occurring before 1995. The study found that negative margins and node metastases were critical prognostic variables, with IHCC having the best prognosis (P < 0.001)[15]. Murakami et al[8] gathered pertinent information from 132 patients with CCA (IHCC: 26, PHCC: 50, DCC: 56) and discovered that negative surgical margins were the most important prognostic markers for patients with CCAs, and that all three forms of CCAs had a comparable prognosis. Both of the aforementioned research failed to adequately highlight the similarities and differences between the three types of CCAs. Consequently, Ercolani et al[17] once more concentrated on the problem among 479 CCA patients (IHCC: 172, PHCC: 243, DCC: 64). 70% of the patients in their group had surgical resections, while the remaining patients received only palliative care. Their findings demonstrated that although there was no discernible difference in OS, R0 resections and microvascular invasions were more commonly found in patients with IHCC. 242 CCA patients (IHCC: 90; PHCC: 104; DCC: 48) were examined by Waseem and Tushar[18], who found that the three types of CCAs differed in their clinical presentation, natural history, and treatment strategy. Patients with DCC had the best prognosis, according to their analysis (P < 0.05). Similar findings were also noted in the Hang et al[20] study, which found that DCC patients had the best prognosis when compared to the other two CCA types (P < 0.001). It's important to note that the Hang et al[20] study was the biggest to date (IHCC: 5462, PHCC: 5626, DCC: 622). It emphasized how target spots differed depending on the location of the tumor, which helped develop tailored adjuvant therapies for CCA patients based on their anatomical characteristics. In addition to the triple-arm studies previously stated, Various two-arm studies were also published to examine the variations in CCAs[19,21,22]. PHCC and ICC had poorer prognoses and pathological characteristics, according to Guglielmi et al's analysis of 33 ICC patients and 62 PHCC patients[19]. After analyzing 236 patients with PHCC and 130 patients with DCC, Nakanishi et al[21] found that their prognoses were comparable. Sallinen et al[22] similarly reported similar findings, stating that patients with DCC had a higher frequency of lymphatic metastases and that survival rates were comparable between those with PHCC and DCC. All of the previously mentioned research were included in our investigation, and their findings and conclusions were further confirmed. Biliary epithelium and peri-biliary glands have been reported to be the source of PHCC and DCC. IHCC, on the other hand, appeared to have a somewhat distinct genesis, with hepatic progenitor cells perhaps serving as an additional source[28]. Their varied origins might explain the changing epidemiology, which shows that the incidence of IHCC has been steadily increasing and that risk factors, including hepatitis B virus infection, have been found to be strongly linked to the development of IHCC[29,30]. According to Sarkhampee et al[24] the sole treatment for CCA is surgical resection; the objective of curative-intent resection is R0 resection since it increases OS over R1 resection, and it provides higher survival rates than palliative resection. Furthermore, CCAs were not frequently found in risk groups, and as previously reported, only a small percentage of patients with CCA were candidates for curative surgery, in contrast to hepatocellular carcinoma, for which a more developed and sophisticated monitoring system has been built[31]. According to Hang et al's study[20], their cohort's time span extended until 1973, which was far older than other published series. Their investigation showed that the resectability rates were 18.5% (IHCC), 37.8% (DCC), and 36.3% (PHCC). Recent decades have seen a significant increase in the resectability rate due to an aggressive surgical program and the development of diagnostic techniques[20]. Ercolani et al[17] found that about 60% of individuals with hilar tumors may have their malignancies removed. Similar alterations have also been documented in several recent data, which show that almost 90% of patients with DCC and 80% of patients with IHCC and PHCC have received curative resections[32,33]. Siripongsakun et al[34] reported encouraging surveillance results using abdomen ultrasonography (HR, 0.41; P = 0.012), indicating the value of using screening methods conventionally in some endemic locations. Furthermore, routine evaluation is advised for people with significant risk factors for CCA, such as sclerosing cholangitis or viral hepatitis. Regarding the possibly viable treatment approaches for CCAs, curative surgery remains the most common method of managing CCA, with adjuvant therapies administered either before to or following surgery. The most significant effect on the survival of patients with resectable illness would be surgical resection, whether or not adjuvant medications are used, regardless of the particular kind of CCA[35]. In contrast to the other two forms of CCAs, patients with IHCC had a much lower incidence of node metastasis and neural invasion, as well as a significantly greater incidence of attaining R0 resection (81%)[17]. According to some investigators, patients with a single tumor that was less than 2 cm in size had a 5-year survival rate of 65% and a 5-year recurrence rate of just 18%, which was better than the outcomes of patients with more advanced IHCC lesions. The prognosis of IHCC was comparable to that of other CCAs in our study and in many other published series, which may be explained by the higher frequency of microvascular invasion found in these individuals[15,16,23]. In some experienced centers, the effectiveness of liver transplantation has been confirmed[33,36,37]. A prospective multi-center clinical study (NCT02878473) is now underway to demonstrate the efficacy of liver transplantation for IHCC. In order to treat PHCC, curative surgery may potentially be used. Nearly 60% of patients with PHCC might have curative surgery, as was previously reported; however, peri-operative mortality and morbidity rates (Clavien grade 3-4) were found to be much higher[17]. It is typically challenging to attain negative surgical margins because of the tumor's infiltrative growth style and peculiar placement, which frequently places it next to nearby critical circulatory systems. The stated postoperative 5-year survival rate is therefore just 20%-40%, even in some very well chosen patients[38,39]. In contrast, some patients who had R0 resections with node metastases were found to have a promising 5-year survival rate of 70%[40,41]. For PHCC, liver transplantation can be the only appealing alternative that doesn't involve significant bile duct spreading. The early-published studies, however, indicated a dismal prognosis with substantial recurrence rates even after the implementation of comprehensive treatment[42,43]. However, as these investigations were retrospective cohort studies without defined and protocolized adjuvant regimens, it is not realistic to exclude the role of liver transplantation in the treatment of PHCC[44]. Undoubtedly, in order to reevaluate the efficacy of liver transplantation for PHCC, major prospective studies are needed in the future. Achieving a negative duct margin is the primary goal of surgically excising DCC since a positive margin is strongly linked to a bad prognosis[45]. Patients with DCC have had pancreaticoduodenectomy, which involves carefully removing the perivascular tissue surrounding the superior mesenteric artery and vein. Also, as we have learned from our study and several other published series[8,15,17,23], Due to the higher frequency of lymphatic invasion in DCC, a regional or more extensive lymphadenectomy should be actively performed. All lymph nodes surrounding the common bile duct and porta hepatitis should be inspected for possible drainage pathways from lower bile duct cancers. It has been shown that nodes surrounding the hepatic artery have prognostic implications in patients with pancreatic ductal carcinoma, and this may also apply to patients with DCC[46].

Notably, our analysis revealed a paradoxical yet consistent finding: DCC demonstrated the most favorable OS despite exhibiting the highest rate of lymph node metastasis (56.9%). This challenges traditional oncological expectations, where extensive lymphatic spread typically correlates with poorer prognosis. One plausible explanation lies in the anatomical and surgical context of DCC. Unlike perihilar or intrahepatic tumors, DCCs are often amenable to pancreaticoduodenectomy, a procedure that enables en bloc resection with extensive regional lymphadenectomy, which may reduce residual disease and improve locoregional control. Moreover, DCCs may become symptomatic earlier due to bile duct obstruction, leading to earlier diagnosis and intervention. These clinical factors, rather than tumor biology alone, may account for the survival advantage observed, despite aggressive pathological features such as nodal involvement.

Limitations

Another important limitation is that while our analysis highlights the paradox of better survival in DCC despite higher lymphatic invasion, the underlying biological mechanisms remain speculative. The available data did not consistently report on the extent of lymphadenectomy, surgical margin status stratified by tumor subtype, or detailed nodal staging beyond simple positivity rates. Additionally, tumor biology such as genomic or immunological differences between subtypes was not assessed in the included studies, limiting our ability to explore potential molecular contributors to these observations.

Future directions

Future research should aim to elucidate the biological and clinical mechanisms underpinning the observed paradox in DCC outcomes. Prospective studies should incorporate standardized reporting of lymph node dissection extent, nodal burden, and recurrence patterns stratified by tumor location. Molecular profiling of DCC, PHCC, and IHCC may help identify subtype-specific drivers of prognosis and treatment response. Moreover, investigation into the tumor microenvironment including immune cell infiltration and stromal features may clarify whether DCC elicits a distinct host response that mitigates the adverse effects of lymphatic invasion. Such studies will be essential to validate and mechanistically explain our findings and support personalized treatment strategies.

CONCLUSION

Our study provides an in-depth analysis among the three types of CCAs: IHCC, PHCC and DCC. We found that DCC shared the best prognosis vs IHCC and PHCC. Moreover, lymphatic invasion is more frequently detected in patients with DCC. DCC shares a more favorable prognosis than IHCC. Future more powerful well-designed studies are warranted for further validation.

ACKNOWLEDGEMENTS

The authors sincerely thank the referenced studies or consortiums for contributing open-access datasets for the analysis.