Neoadjuvant therapy use and association with postoperative outcomes and overall survival in patients with extrahepatic cholangiocarcinoma

Abstract

Background and Objectives:

Evidence for neoadjuvant therapy (NAT) in extra-hepatic cholangiocarcinoma (eCCA) is limited. Our objectives were to: (1) characterize treatment trends, (2) identify factors associated with receipt of NAT, and (3) evaluate associations between NAT and postoperative outcomes.

Methods:

Retrospective cohort study of the National Cancer Database (2004–2017). Multivariable logistic regression assessed associations between NAT and postoperative outcomes. Stratified analysis evaluated differences between surgery first, neoadjuvant chemotherapy, and neoadjuvant chemoradiation (CRT).

Results:

Among 8040 patients, 417 (5.2%) received NAT. NAT increased during the study period 2.9%–8.4% (p < 0.001). Factors associated with receipt of NAT included age <50 (vs. >75, odds ratio [OR] 4.32, p < 0.001) and stage 3 disease (vs. 1, OR 1.68, p = 0.01). Compared with surgery first, patients who received NAT had higher odds of R0 resection (OR 1.49, p = 0.01) and lower 30-day mortality (OR 0.51, p = 0.04). On stratified analysis, neoadjuvant chemotherapy was not associated with differences in any outcomes. However, neoadjuvant CRT was associated with improvement in R0 resection (OR 3.52, <0.001) and median survival (47.8 vs. 25.3 months, log-rank < 0.001) compared to surgery first.

Conclusions:

NAT, particularly neoadjuvant CRT, was associated with improved postoperative outcomes. These data suggest expanding the use of neoadjuvant CRT for eCCA.

1 |. INTRODUCTION

Cholangiocarcinoma (CCA) is the most common malignancy of the biliary tract, and CCA accounts for approximately 20% of deaths from hepatobiliary cancers worldwide.¹ The majority of patients present with advanced disease, and biliary obstruction and vascular compromise to the liver are significant risks for morbidity and mortality.^2,3 Extrahepatic cholangiocarcinoma (eCCA) in particular presents many management challenges. Upfront surgery is potentially curative and considered standard of care for those with resectable disease, though the operations are extensive, aborted procedures due to unresectable disease are not uncommon, and 5-year survival remains poor.^4–6 Neoadjuvant therapy (NAT) represents a potential management strategy. NAT has been increasingly used in other biliary and pancreatic neoplasms, as long postoperative recovery periods and complications often prevent the administration of adjuvant therapy. In the treatment of this aggressive disease, NAT also aids in selection of patients who are ultimately likely to benefit from surgery.^7,8 While many patients with eCCA receive adjuvant therapy, particularly in the case of positive surgical margins or lymph node involvement, the use of chemotherapy and/or radiation therapy in the preoperative setting is less common.⁹ Some studies have suggested improved surgical resectability and operative margin status with the use of neoadjuvant chemoradiation.^10,11 However, these studies were small and thus the potential role of NAT is not well understood.

Little is known about the current practices in the use of NAT or the association between receipt of NAT and postoperative outcomes and overall survival (OS). The objectives of this study were to (1) characterize treatment trends over time, (2) identify patient and hospital factors associated with the use of NAT, and (3) evaluate the association between NAT and short- and long-term outcomes.

2 |. MATERIALS AND METHODS

2.1 |. Data source

The National Cancer Database (NCDB) is a nationwide registry sponsored by the American College of Surgeons, the Commission on Cancer, and the American Cancer Society. The NCDB includes data from more than 1500 facilities across the United States and represents more than 70% of newly diagnosed malignancies.¹² Details of data abstraction, sampling strategy, and variables collected are described in the literature.^13,14 This study was deemed exempt by the Institutional Review Board of Northwestern University because it used pre-existing, deidentified data.

2.2 |. Study population

For this retrospective cohort study, data from the NCDB were used to identify patients who underwent surgical resection of eCCA (International Classification of Diseases for Oncology code C24.0 and histology codes 8140, 8160, 8162, 8144, 8310, 8481, and 8260) from January 1, 2004 to December 31, 2017. Distal common bile duct cancers were not included. Patients were excluded if they had metastatic disease at the time of diagnosis, if they were diagnosed and treated at different facilities, or if receipt and/or sequence of surgery, chemotherapy, or radiation was unknown. Patients were classified as having received NAT if they underwent neoadjuvant chemotherapy or chemoradiation. Patients who received neoadjuvant radiation alone were excluded from the analysis. The remaining patients were classified as surgery first. Inclusion and exclusion criteria are summarized in Figure 1.

FIGURE 1.

Patient selection schema. CRT, chemoradiation; NCDB, National Cancer Database.

2.3 |. Outcomes

This study evaluated both postoperative and long-term outcomes. Postoperative outcomes included surgical R0 resection, unplanned readmission within 30 days postoperatively, 30-day mortality, and 90-day mortality. Long-term outcomes included OS.

2.4 |. Covariates

Covariates included demographic, socioeconomic, and clinical variables in the NCDB. The following were considered for this study: age, sex, race (White, Black, Hispanic, Asian, Other/Unknown), income, insurance type, Charlson-Deyo score, distance from hospital, tumor differentiation and stage, year of diagnosis, and hospital type, location, and surgical case volume. Insurance status was characterized as private, government (Medicare or Medicaid), uninsured, and other/unknown. Number of coexisting medical conditions was captured with the Charlson–Deyo score (0, 1, and 2+). Clinical stage of disease was determined using the American Joint Committee on Cancer (AJCC) staging system and was used in the analysis of receipt of NAT, as it was available to physicians preoperatively when decisions regarding management are made. AJCC pathologic stage, which was more often complete, was used in the analysis of OS. Types of surgery undergone by patients included standard or radical resection.

2.5 |. Statistical analysis

Proportional trends in practice patterns were assessed using the Cochran-Armitage test for trend. Baseline patient, oncologic, and hospital characteristics were evaluated using χ² tests and t-tests, as appropriate. All tests of statistical significance were two-sided with α = 0.05. Multivariable logistic regression models with robust standard errors were constructed to determine independent predictors of receipt of NAT while accounting for patient clustering within hospitals. Variables were included in the model if they were clinically relevant or found to be significant with α = 0.05 in bivariate analysis. Next, we evaluated the association of NAT with postoperative outcomes. Multivariable logistic regression models were adjusted by propensity score (PS) to account for potential selection bias in the receipt of NAT. PS was calculated from a separate logistic regression model predicting the receipt of NAT while controlling for age, tumor grade and stage, and year of diagnosis using the Stata package psmatch2. We first compared differences in outcomes between Surgery First and NAT groups. Given that NAT included both neoadjuvant chemotherapy and CRT, we then conducted a stratified analysis to evaluate differences between Surgery First, Neoadjuvant Chemotherapy, and Neoadjuvant CRT groups. Those who eventually received adjuvant therapy were not considered separately in these analyses because the events considered would have likely preceded receipt of adjuvant therapy. All analyses were performed with robust standard errors to account for clustering at the hospital level.

Survival analysis first compared differences between Surgery First and NAT groups, followed by a stratified analysis comparing Surgery First, Neoadjuvant Chemotherapy, and Neoadjuvant CRT. Differences between Surgery Alone, Adjuvant Therapy (chemotherapy or CRT), and NAT (chemotherapy or CRT) were then assessed to account for adjuvant therapy as a potential confounder of survival differences. Subjects were dropped from this analysis if survival data was incomplete. We used the Kaplan–Meier estimator of survival to plot OS and used the log-rank test to compare differences in survival across groups. Cox proportional-hazard models were used to estimate the risk of death due to any cause.

Risk-adjusted associations, accounting for clustering at the hospital level using robust standard errors, are presented as odds ratios (ORs) or hazard ratios (HRs) with 95% confidence intervals (CIs). All statistical analyses were performed in Stata SE version 17.0 (StataCorp LLC).

3 |. RESULTS

3.1 |. Patient characteristics

Among 8040 patients with eCCA who underwent surgery in 910 hospitals during the study period, 417 (5.2%) received NAT. Of those who received NAT, 215 (51.6%) received neoadjuvant chemotherapy and 202 (48.4%) received neoadjuvant CRT. Patients who received neoadjuvant CRT were more likely to undergo radical resection (48.5%) compared to those who received surgery first (33.9%) or neoadjuvant chemotherapy (30.7%, p < 0.001). Of the 7623 patients who underwent upfront surgery, 4057 (50.5%) received adjuvant therapy (chemotherapy: 1588 (39.1%), CRT: 2469 (60.9%)). Patient and hospital characteristics are summarized in Table 1.

TABLE 1.

Characteristics of patients who underwent surgery for extrahepatic cholangiocarcinoma

	Surgery first (n = 7623 [94.8%])	Neoadjuvant therapy (n = 417 [5.2%])	p Value
Age, median (IQR), years	68 (60–75)	63 (55–70)	<0.001
Sex			0.49
Male	4826 (63.3)	257 (61.6)
Female	2797 (36.7)	160 (38.4)
Race/ethnicity			0.002
White	5938 (77.9)	360 (86.3)
Non-Hispanic Black	580 (7.6)	21 (5.0)
Hispanic	529 (6.9)	17 (4.1)
Asian	320 (4.2)	9 (2.2)
Other	256 (3.4)	10 (2.4)
Location			<0.001
New England	462 (6.1)	14 (3.4)
Mid Atlantic	1455 (19.1)	51 (12.2)
South Atlantic	1526 (20.0)	70 (16.8)
East North Central	1357 (17.8)	69 (16.6)
East South Central	370 (4.9)	8 (1.9)
West North Central	506 (6.6)	65 (15.6)
West South Central	665 (8.7)	29 (7.0)
Mountain	276 (3.6)	49 (11.8)
Pacific	892 (11.7	32 (7.7)
Unknown	114 (1.5)	30 (7.2)
Income			<0.001
<$40 227	1108 (14.5)	43 (10.3)
$40 227-$50 353	1503 (19.7)	82 (19.7)
$50 354-$63 332	1651 (21.7)	80 (19.2)
>$63 333	2756 (36.2)	146 (35.0)
Unknown	605 (7.9)	66 (15.8)
Insurance status			<0.001
Private	2771 (36.4)	187 (44.8)
Medicaid/Medicare	4409 (57.8)	202 (48.4)
Uninsured	197 (2.6)	7 (1.7)
Other	246 (3.2)	21 (5.0)
Charlson-Deyo			0.08
0	5386 (70.7)	314 (75.3)
1	1623 (21.3)	70 (16.8)
2+	614 (8.0)	33 (7.9)
Clinical Stagea			0.01
1	1430 (40.6)	112 (36.8)
2	1540 (43.7)	125 (41.1)
3	551 (15.7)	67 (22.0)
Type of surgery			0.08
Standard resection	4667 (61.2)	235 (56.4)
Radical resection	2586 (33.9)	164 (39.3)
Not specified	370 (4.9)	18 (4.3)
Distance from treatment facility			<0.001
0–49 miles	5332 (70.0)	218 (52.3)
50–99 miles	947 (12.4)	57 (13.7)
100+ miles	1344 (17.6)	142 (34.1)
Hospital type			<0.001
Academic	4573 (60.0)	261 (62.6)
Nonacademic	2936 (38.5)	126 (30.2)
Unknown	114 (1.5)	30 (7.2)
Surgical cases/facility			<0.001
1st quartile	2025 (26.6)	89 (21.3)
2nd quartile	1846 (24.2)	80 (19.2)
3rd quartile	1911 (25.1)	95 (22.8)
4th quartile	1841 (24.1)	153 (36.7)
Year of diagnosis			<0.001
2004–2005	775 (10.2)	23 (5.5)
2006–2007	878 (11.5)	33 (7.9)
2008–2009	974 (12.8)	37 (8.9)
2010–2011	1122 (14.7)	46 (11.0)
2012–2013	1223 (16.0)	62 (14.9)
2014–2015	1301 (17.1)	92 (22.1)
2016–2017	1350 (17.7)	124 (29.7)

^aClinical stage of disease determined using the American Joint Committee on Cancer (AJCC) staging system.

3.2 |. Trends in use of neoadjuvant therapy

Overall, the proportion of patients undergoing surgery alone decreased during the study period from 52.1% to 35.1% (p < 0.001). Use of adjuvant therapy increased 45.0%–56.6%, while management with NAT increased during the study period from 2.9% to 8.4% (p < 0.001, Figure 2). Use of neoadjuvant chemotherapy increased from 0.5% to 5.8%, while approximately 2% of patients received neoadjuvant CRT throughout the study period.

FIGURE 2.

Percentage of patients receiving adjuvant therapy and neoadjuvant therapy (NAT) over time

3.3 |. Predictors of receipt of neoadjuvant therapy

Compared with surgery first, patients who received NAT were younger, lived further from the treatment facility, and were of higher clinical stage at time of diagnosis (Table 1). These associations were also present on multivariable analysis, which showed age <50 years (vs. >75 years, OR 4.31, 95% CI 2.55–7.29), stage 3 disease (vs. stage 1, OR 1.68, 95% CI 1.12–2.52), distance traveled >100 miles (vs <50 miles, OR 2.28, 95% CI 1.72–3.04), and diagnosis in 2016–2017 (vs. 2004–2005, OR 2.37, 95% CI 1.15–4.86) were independently associated with NAT use (Table 2).

TABLE 2.

Factors associated with the receipt of neoadjuvant therapy

Characteristic	OR (95% CI)	p Value
Age
>75	Reference
50–74	1.60 (1.08–2.37)	0.02
<50	4.32 (2.55–7.29)	<0.001
Race/ethnicity
Non-Hispanic White	Reference
Non-Hispanic Black	0.63 (0.37–1.08)	0.09
Hispanic	0.69 (0.37–1.30)	0.25
Asian	0.61 (0.26–1.44)	0.26
Other/unknown	0.53 (0.25–1.15)	0.11
Insurance status
Private	Reference
Government	1.00 (0.75–1.33)	0.98
Uninsured	0.80 (0.34–1.90)	0.62
Other/unknown	1.54 (0.79–3.01)	0.20
Charlson-Deyo
0	Reference
1	0.75 (0.52–1.06)	0.10
2+	0.92 (0.60–1.41)	0.69
Clinical Stagea
1	Reference
2	1.12 (0.78–1.62)	0.54
3	1.68 (1.12–2.52)	0.01
Distance
0–49 miles	Reference
50–99 miles	1.25 (0.90–1.75)	0.84
100+ miles	2.28 (1.72–3.04)	<0.001
Surgical case/facility
1st quartile	Reference
2nd quartile	1.08 (0.71–1.63)	0.72
3rd quartile	0.76 (0.46–1.25)	0.28
4th quartile	1.42 (0.91–2.22)	0.12
Year of diagnosis
2004–2005	Reference
2006–2007	1.05 (0.54–2.06)	0.88
2008–2009	0.78 (0.35–1.72)	0.53
2010–2011	0.95 (0.44–2.03)	0.89
2012–2013	1.13 (0.48–2.68)	0.78
2014–2015	1.80 (0.78–4.14)	0.17
2016–2017	2.37 (1.15–4.86)	0.02

Abbreviations: CI, confidence interval; OR, odds ratio.

^aClinical stage of disease was determined using the American Joint Committee on Cancer (AJCC) staging system.

3.4 |. Association between NAT and postoperative outcomes

Compared with surgery first, patients who received NAT had higher odds of R0 resection (OR 1.49, 95% CI 1.10–2.02), lower 30-day mortality (OR 0.51, 95% CI 0.27–0.97), and lower 90-day mortality (OR 0.58, 95% CI 0.35–0.97), controlling for age, tumor differentiation and stage, type of surgical procedure performed, surgical volume, year of diagnosis, and estimated probability of NAT use. There were no significant differences in unplanned 30-day readmission between NAT and Surgery First groups (Table 3). On stratified analysis, neoadjuvant chemotherapy was not associated with differences in any of these outcomes compared to surgery first. However, neoadjuvant CRT was associated with improvement in R0 resection (OR 3.52, 95% CI 2.11–5.86) and 90-day mortality (OR 0.33, 95% CI 0.13–0.81) compared to surgery first (Table 3). Radical resection was also independently associated with increased odds of R0 resection compared to standard resection (OR 1.39, 95% CI 1.23–1.57) and was not significantly associated with differences in readmission, 30-day mortality, or 90-day mortality.

TABLE 3.

Association between receipt of NAT and postoperative outcomes

	OR (95% CI)	p Value
R0 resectiona,b
Overall NAT	1.49 (1.10–2.02)	0.01
Neoadjuvant chemotherapy	0.95 (0.68–1.33)	0.75
Neoadjuvant CRT	3.52 (2.11–5.86)	<0.001
30-day readmissiona,b
Overall NAT	0.98 (0.67–1.43)	0.90
Neoadjuvant chemotherapy	0.84 (0.51–1.38)	0.49
Neoadjuvant CRT	1.15 (0.70–1.91)	0.58
30-day mortalitya,b
Overall NAT	0.51 (0.27–0.97)	0.04
Neoadjuvant chemotherapy	0.53 (0.23–1.25)	0.15
Neoadjuvant CRT	0.49 (0.18–1.34)	0.16
90-day mortalitya,b
Overall NAT	0.58 (0.35–0.97)	0.04
Neoadjuvant chemotherapy	0.83 (0.44–1.54)	0.55
Neoadjuvant CRT	0.33 (0.13–0.81)	0.02

Abbreviations: CI, confidence interval; CRT, chemoradiation; NAT, neoadjuvant therapy; OR, odds ratio.

^aControlling for age, differentiation, stage, surgical volume, surgical procedure, and year of diagnosis.

^bEach outcome modeled separately. Surgery First versus NAT and Surgery First versus Neoadjuvant Chemotherapy versus Neoadjuvant CRT also represent separate models. Reference category for each model is Surgery First.

3.5 |. Survival analysis

NAT was associated with a longer median survival time compared to surgery first (35.1 vs. 25.3 months, log-rank < 0.001), and on stratified analysis neoadjuvant CRT was associated with the best median survival compared to both neoadjuvant chemotherapy and surgery first (47.8 vs. 30.3 vs. 25.3 months, log-rank < 0.001). On subsequent analysis comparing adjuvant and neoadjuvant strategies to surgery alone, neoadjuvant CRT demonstrated the longest median survival time (47.8 months), while neoadjuvant chemotherapy was associated with a modest increase in median survival compared to adjuvant chemotherapy and adjuvant CRT (median 30.3 vs. 28.5 vs. 28.7 months, log-rank < 0.001, Figure 3). All systemic and/or locoregional regimens were associated with improved OS compared to surgery alone (Figure 3). On multivariable analysis, neoadjuvant CRT was associated with the best improvement in OS (HR 0.64, 95% CI 0.52–0.79) compared to surgery alone, though neoadjuvant chemotherapy and both forms of adjuvant therapy were also significantly associated with improved OS (Supporting Information: Table 1).

FIGURE 3.

Survival differences associated with the use of neoadjuvant therapy in patients who underwent surgical resection for extrahepatic cholangiocarcinoma. CRT, chemoradiation.

4 |. DISCUSSION

Outcomes for eCCA are poor, and while NAT represents a potential management strategy, little is known about its use or effectiveness. In this study, a national cohort of patients with nonmetastatic disease was analyzed to ascertain factors and outcomes associated with the use of NAT. Several factors such as younger age, advanced clinical stage, and increased distance from treatment facility were associated with use of NAT. Patients who received NAT were more likely to experience R0 resection and had improved OS compared to those who received surgery first. Notably, these differences appear to be primarily driven by improved outcomes with the receipt of neoadjuvant CRT rather than neoadjuvant chemotherapy alone.

Though NAT has been increasingly used in the management of intrahepatic CCA and pancreatic adenocarcinoma, practice patterns in eCCA are less well described.^15,16 In our national cohort, use of NAT increased over the study period, with 2.8% of patients in 2004-–2005 and 8.4% of patients in 2016–2017 receiving neoadjuvant chemotherapy or CRT. Use of neoadjuvant chemotherapy in particular increased over the study period while use of neoadjuvant CRT remained stable. This increased use of NAT for eCCA is consistent with that seen in other malignancies and is potentially attributable to evolving chemotherapy regimens and a developing body of outcomes research.

There are many theoretical benefits to NAT, including treatment of occult systemic disease and downstaging tumors to improve resectability and odds of resection with negative margins.^17,18 NAT also helps with surgical patient selection; it identifies those with aggressive tumors who progress to overt metastatic disease and avoids surgery in these patients who would be unlikely to benefit from operative intervention. Though upfront surgery has long been considered standard of care for eCCA, current National Comprehensive Cancer Network (NCCN) guidelines emphasize a paucity of data to support NAT and encourage clinical trial participation.¹⁹ Indeed, some prospective clinical trials are currently investigating the potential role of NAT, though these trials are not expected to conclude for several years. Also, of note, none of these trials include neoadjuvant radiation.^20–23 As these trials continue, our observational study provides pragmatic, real-world data on patients treated with NAT.

In this retrospective study, patients with stage III disease were at increased odds of receiving NAT relative to those with stage 0 or stage I disease, suggesting that NAT is used more frequently for those with advanced disease. While guidelines for other malignancies, such as pancreatic adenocarcinoma, distinguish between “resectable” and “borderline resectable” tumors based on location and local invasion, those for eCCA make no such distinction. Our data suggest that current practices may favor the use of NAT for similarly advanced disease. It is important to note that patients of lower clinical stage often receive adjuvant therapy and are included in clinical trials and therefore may also benefit from NAT.

Some of the most striking findings of this study are those involving neoadjuvant CRT. While some prior studies have investigated outcomes of neoadjuvant chemotherapy, very few have studied neoadjuvant CRT, and results have been mixed.^9,10,24,25 Katayose et al. conducted a clinical trial of neoadjuvant CRT with 25 patients with eCCA and found an R0 resection rate of 90%.²⁶ However, this study was quite small and lacked a control group, limiting the generalizability of the findings. A retrospective review of 37 patients did not show a difference in outcomes between patients who received neoadjuvant chemotherapy and those who received neoadjuvant CRT, though multivariable analysis was not performed.²⁷ In our large database study, neoadjuvant CRT was associated with improved odds of margin-negative resection and lower 90-day mortality, while neoadjuvant chemotherapy was not associated with improvements in any short-term outcomes relative to a surgery first approach. While neoadjuvant chemotherapy treats possible micro-metastases, neoadjuvant CRT is more likely to shrink tumors, making the technically challenging resection more feasible, and helping obtain local control.

Some recent clinical trials on NAT in eCCA have yet to report survival data, while others have lacked statistical power to demonstrate improved survival.^28–30 Yadav et al. conducted a retrospective analysis of patients with CCA and found longer OS for patients who received NAT compared to those who underwent upfront surgery followed by adjuvant therapy. However, this study was limited to neoadjuvant chemotherapy.³¹ In this study, we show neoadjuvant CRT to have the best OS when compared with surgery first, neoadjuvant chemotherapy, and adjuvant chemotherapy and CRT. One rationale for the use of NAT is that postoperative complications and recovery may limit or delay the administration of adjuvant therapy, thus precluding the survival benefits of chemoradiation.^7,8 While our study does show the shortest median survival time for those who are treated with surgery alone, we also see a longer median survival for neoadjuvant CRT relative to adjuvant CRT, suggesting a benefit to receiving CRT in the preoperative setting. Improvements in OS with receipt of CRT may be related in part to surgical procedure and R0 resection. In this study cohort, a higher percentage of patients who received neoadjuvant CRT underwent radical resection than those who received surgery first or neoadjuvant chemotherapy. Other studies have established R1 resection as a poor prognostic indicator in CCA, and in this study, patients who received neoadjuvant CRT had increased odds of R0 resection, which likely contributed to the observed OS benefit.^32,33

5 |. LIMITATIONS

There are limitations to this study. Many of these are inherent to the use of the NCDB: data are retrospective, and we may not have accounted for all potential confounders.¹⁴ The NCDB does not contain data on chemotherapy regimens, which may influence postoperative outcomes and OS. It similarly does not report data on complications of chemotherapy or radiotherapy. Preoperative management of eCCA often includes biliary drainage, though the safety of NAT in this setting is not well understood. There is also possible selection bias, as we do not include patients who received NAT but do not make it to surgery due to disease progression. However, one of the benefits of NAT is that it facilitates the selection of patients who may benefit from surgical resection and avoids intervention, and potential morbidity and complication, in those who would not.

Neoadjuvant CRT remains a rarely used management strategy for patients with eCCA, accounting for just 2% of patients in 2016–2017. Data on outcomes with the receipt of NAT is scarce, and upfront surgery remains standard of care for those with resectable disease. However, our study shows an association between neoadjuvant CRT and R0 resection, lower 90-day mortality, and improved OS, offering compelling evidence supporting expanding the use of neoadjuvant CRT for eCCA.

6 |. CONCLUSIONS

The use of NAT for eCCA increased over the study period but remained uncommonly used, and was associated with several factors, including advanced-stage disease. NAT, particularly neoadjuvant CRT, was associated with improved R0 resection, lower 90-day mortality, and longer OS compared to an upfront surgery approach. Data from ongoing prospective clinical trials may yield further evidence on the relative benefit of NAT relative to upfront surgery in eCCA.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the NCDB. Data are available from the corresponding author upon request, though restrictions may apply per the NCDB.