Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Distant Metastases in Patients with Intrahepatic Cholangiocarcinoma: Does Location Matter? A Retrospective Analysis of 370 Patients

Distant Metastases in Patients with Intrahepatic Cholangiocarcinoma: Does Location Matter? A... Hindawi Journal of Oncology Volume 2020, Article ID 7195373, 8 pages https://doi.org/10.1155/2020/7195373 Research Article Distant Metastases in Patients with Intrahepatic Cholangiocarcinoma: Does Location Matter? A Retrospective Analysis of 370 Patients 1 1 1 2 Felix Hahn , Lukas Mu¨ller, Aline Ma¨hringer-Kunz, Yasemin Tanyildizi, 3 1 4 4,5 Daniel Pinto dos Santos, Christoph Du¨ber, Peter R. Galle, Arndt Weinmann, and Roman Kloeckner Department of Diagnostic and Interventional Radiology, University Medical Center of the Johannes Gutenberg-University Mainz, Mainz, Germany Department of Neuroradiology, University Medical Center of the Johannes Gutenberg-University Mainz, Mainz, Germany Department of Radiology, University Hospital of Cologne, Cologne, Germany Department of Internal Medicine, University Medical Center of the Johannes Gutenberg-University Mainz, Mainz, Germany Clinical Registry Unit (CRU), University Medical Center of the Johannes Gutenberg-University Mainz, Mainz, Germany Correspondence should be addressed to Felix Hahn; felix.hahn@unimedizin-mainz.de Received 18 May 2020; Revised 13 August 2020; Accepted 28 September 2020; Published 10 October 2020 Academic Editor: Dali Zheng Copyright © 2020 Felix Hahn et al. ,is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background. Intrahepatic cholangiocarcinoma (ICC) is an aggressive tumor entity, and distant metastases are common. However, studies investigating patterns and clinical relevance of distant metastases are rare. ,erefore, we aimed to analyze occurrence, location, and prognostic impact of distant metastases on overall survival (OS). Methods. Between 1997 and 2018, 417 patients with ICC were treated at our tertiary care center. Distant metastases and intrahepatic tumor burden were retrospectively evaluated in a longitudinal approach using volumetric assessment of cross-sectional imaging studies and all available medical/histopathological reports. Results. Finally, 370 patients with histopathologically confirmed ICC were included. Of these, 186 showed distant metastases, either initially (n � 59) or during follow-up (n � 127). ,e most common metastatic sites were the lung (n � 105), peritoneum (n � 81), and bone (n � 50). After detection of lung metastases, the residual median OS was 5.3 months; followed by peritoneal metastases, 4.5 months, and bone metastases, 4.4 months (P � 0.17). At the time of first metastatic occurrence, residual OS according to intrahepatic tumor burden of <25%, 25–50%, and >50% was 6.5 months, 4.9 months, and 1.2 months, re- spectively (P< 0.001). In multivariate hazard regression, hepatic tumor burden, liver function, and subsequent treatment were significant predictors of survival. Conclusions. During the disease course, every second patient developed extrahepatic metastases. While the presence of distant metastases was associated with poor patient outcomes, there was no significant difference between metastatic sites. However, hepatic tumor burden was the life-limiting risk factor in a majority of patients at the time of distant metastatic disease. Affected patients are often asymptomatic in the early 1. Introduction stages; thus at diagnosis, the tumor is often already at an Intrahepatic cholangiocarcinoma (ICC) is the second most advanced stage [5]. Nonresectable patients have repeatedly common primary liver malignancy after hepatocellular been identified as having a poorer prognosis than patients carcinoma (HCC). Incidence in Western countries is esti- undergoing surgery [6, 7]. However, even if resection is mated to be approximately 0.4–2.0/100,000 and has con- possible, tumor recurrence has been reported in about 65% siderably increased since the last three decades [1–4]. of the patients [8]. 2 Journal of Oncology Regarding systemic chemotherapy, the combination of gemcitabine and cisplatin has been widely administered as first-line chemotherapy since the publication of the multi- center UK-ABC 02 study in 2010 [7]. Even though systemic chemotherapy is the mainstay of treatment, intra-arterial therapies such as transarterial chemoembolization (TACE) or selective internal radiation therapy (SIRT) have also been established since the last decade as treatment options for selected patients [9]. However, prognosis remains poor [10]. Figure 1: Exemplary contrast-enhanced axial CTslices in the portal venous phase depicting the volumetric measurement of tumor-free As resection is the only curative treatment option, the liver tissue for one patient using Aquarius iNtuition software. presence of distant metastases impedes curative-intent re- sections [11]. Distant metastases are incorporated into the prognostic Wang nomogram [12]; also, the discriminator th 25–50%, >50%). Patients after resection without hepatic between stage III and stage IV in the current 8 edition of recurrence but distant metastases were categorized as “ex- the UICC staging system is the presence of distant metas- clusively extrahepatic metastases” (EEM). tases [13]. In a recent study performed by our group, distant Occurrence and distribution of distant metastases metastases were independent predictors of poor survival as were evaluated using the statistical software package R well [14]. 3.5.1 [16]. Survival analyses were performed using the Based on the Surveillance, Epidemiology, and End Re- “survival” and “survminer” (https://cran.r-project.org/ sults (SEER) database, Wu et al. [15] investigated patterns of package�survival, https://cran.r-project.org/ distant extrahepatic metastases in primary liver cancer, both package�survminer, accessed on 31.12.2019) packages. HCC and ICC. However, the authors limited their inves- In rare cases of missing laboratory values, these were tigation to patients with distant metastases at the time of imputed using the “mice” package (https://cran.r-project. diagnosis only. Moreover, survival analyses were presented org/package�mice, accessed on 31.12.2019). Log-rank for pooled HCC and ICC patients, with ICC patients ac- tests and Kaplan–Meier curves were used for survival counting for only 11% of the study population. analysis between strata. Univariate and multivariate Cox To the best of our knowledge, no study has focused on proportional hazards regression models were fitted to the longitudinal investigation of metastatic occurrence and determine the influence of predictors. As this analysis has the impact of different sites of metastatic spread for patients exploratory intention, P-values should be interpreted in a with ICC. ,erefore, the aim of this study was to analyze descriptive manner. P values< 0.05 were considered occurrence, location, and prognostic impact of distant statistically significant. metastases on overall survival (OS) during the course of disease. 3. Results 2. Materials and Methods 3.1. Occurrence and Location of Distant Metastases. Of the Between January 1997 and January 2018, 417 patients with 417 patients, 47 had to be excluded for reasons as de- histopathologically confirmed ICC were treated at our scribed in the STROBE flowchart (Figure 2); the tertiary care center. ,ese patients were retrospectively remaining 370 patients were included in this study. A identified from a dedicated, prospectively populated clinical total of 200/370 patients underwent primary surgical database. Data from follow-up visits were extracted from the resection, while 170/370 patients were considered non- hospital and radiology information systems. Death dates resectable by an interdisciplinary board due to their were queried at the appropriate resident’s registration of- general state of health or advanced tumor stage asserted fices. Follow-up was ended on December 31, 2018. In case of by imaging studies or surgical exploration. Of the 200 loss to follow-up, patients were censored at the date of last patients undergoing resection, 7 had distant metastases at contact. ,e study was approved by the responsible ethics initial diagnosis, while 68 developed distant metastases committee for the retrospective analysis of clinical data during follow-up. Of the 170 nonresectable patients, 52 (Permit No. 2018–13618). initially presented with distant metastases and 59 de- We evaluated contrast-enhanced computed tomography veloped distant metastases during the course of disease. (CT) or magnetic resonance imaging at diagnosis and during Further baseline characteristics of the patients at the time the patients’ course of disease to determine the size and of the first development of distant metastases are depicted number of intrahepatic lesions as well as the presence of in Table 1. distant metastases in the lung, the peritoneum, bones, other ,e most common metastatic site was the lung (n � 105, organs, and soft tissues other than lymph nodes. At the time 28%), followed by the peritoneum (n � 81, 22%) and bones of first development of distant metastases, hepatic tumor (n � 50, 14%). Other metastatic sites included the adrenals burden as percentage of total liver parenchyma was mea- (n � 8, 2%), brain (n � 3, 1%), spleen (n � 5, 1%), and soft sured by volumetric assessment using dedicated third-party tissues/skin (cutaneous metastases) (n � 8, 2%); these sites software (Aquarius iNtuition ; TeraRecon, Foster City, CA, are henceforth referred to as “other.” ,e time distribution USA; Figure 1) and categorized into three groups (<25%, of metastatic occurrence for each site is depicted in Figure 3. Journal of Oncology 3 recurrence (Figure S2, median residual OS 6.4 months vs. Patients with 20.6 months, P � 0.01). histopathologically In multivariate Cox hazard regression, hepatic tumor confirmed ICC (n = 417) burden, liver function, and subsequent treatment were significant predictors of survival (Figure 6). While the Cross-sectional imaging categories of “bone” and “other” metastases showed an at initial diagnosis missing (n = 36) increased hazard, the effect was not significant. Hepatic tumor burden was the factor with the highest hazard ratios (HRs) in multivariate analysis. Lost to follow-up early after discharge (n = 6) 4. Discussion Cross-sectional imaging In our cohort, the number of patients with extrahepatic at time of metastatic metastases was high with every second patient developing disease missing (n = 5) extrahepatic metastases during the course of disease. Distant metastases were associated with poor outcome, irrespective Finally included of the metastatic site. ,e rarer metastatic locations showed (n = 370) an increased hazard; however, the effect was not significant. In contrast, the amount of hepatic tumor burden and the patient’s ability to tolerate further treatment were the strongest predictors of survival in multivariate Cox hazard Metastatic disease No metastases in regression analysis. (n = 186) follow-up (n = 184) Literature on the investigated topic is scarce, particularly Figure 2: STROBE flowchart showing the number and the reasons regarding the longitudinal investigation of metastatic oc- for dropout. currence and different sites of metastatic spread. ,e SEER- based study by Wu et al. [15] investigated patterns of distant extrahepatic metastases at initial diagnosis in both HCC and 3.2. Survival of Patients with Distant Metastases. When in- ICC. However, they only present pooled survival analyses for vestigating survival from the time of detection of metastatic both cancer entities. Moreover, while the authors claim in spread until death or loss to follow-up, all metastatic sites their conclusion that there were profound differences in risk showed similar survival curves without any statistically of mortality among distant extrahepatic metastatic sites, the significant difference. After detection of lung metastases, the presented confidence intervals of HRs overlap for lung, median OS was 5.3 months; after peritoneal metastases, it bone, and distant lymph node metastases, with only the HR was 4.5 months; and after bone metastases, it was 4.4 of brain metastases significantly differing from the other months. Pooled patients with metastases other than the three metastatic sites. aforementioned locations showed an OS of 4.5 months In both the study by Wu et al. [15] and the current study, (Figure 4). Pairwise comparisons between survival times lung metastases were the most common extrahepatic manifestation. However, in their study, the second most using the log-rank test resulted in P values ranging from 0.29 to 0.75, with the lowest P value of 0.29 between patients with common metastases were bone metastases. It is unclear lung metastases and those with bone metastases. whether peritoneal metastases, the second most common However, when stratifying patients according to hepatic metastases in our study, were investigated. Frega et al. [17] tumor burden at the time of first metastatic occurrence, investigated the occurrence of brain metastases from biliary survival was significantly different. ,e residual survival of tract cancer in 450 patients and found an incidence of about formerly resected patients with exclusively extrahepatic 1.4%, with a median OS from detection of brain metastases metastases was 16.3 months compared with 6.5 months, 4.9 of 4 months. ,is corresponds with the subset of patients months, and 1.2 months for patients with <25%, 25–50%, developing brain metastases in our cohort. Older, autopsy- and >50% hepatic tumor burden, respectively (Figure 5). based studies featured only small patient numbers not ex- Pairwise comparisons among survival times resulted in P ceeding 50 patients, and the share of patients affected by values< 0.01 for all combinations except comparing the distant metastases differs considerably among studies. Lung “<25%” with the “25–50%” group, where the P value was metastases have been reported in 20–40%, and bone me- 0.12. tastases in 10–55% of patients [18–20]. ,e results presented Kaplan–Meier curves of patients with metastatic in our study fall within these broad ranges. disease at initial diagnosis compared with patients Even though the survival of patients with ICC was without metastatic spread at initial diagnosis can be generally poor once metastases were detected, there was no found in Supplementary Materials (Figure S1, median OS discernible influence of metastatic sites on survival in our 4.2 months vs. 17.7 months, P< 0.001). Moreover, sur- cohort. ,is corresponds with clinical routine in that liver vival analysis was performed comparing patients with failure, and especially, biliary complications can pose life- metastatic disease at the time of tumor recurrence with threatening situations for patients with advanced ICC [21], patients without metastatic disease at the time of whereas, for example, respiratory failure due to lung 4 Journal of Oncology Table 1: Baseline characteristics of patients with distant metastases. Initially resected Initially nonresectable All (n � 186) (n � 75) (n � 111) Age, years, median (IQR) 63.8 (56–72) 61.3 (55–68) 67.0 (57–74) Male 112 (60.2) 43 (57.3) 69 (62.2) Sex, n (%) Female 74 (39.8) 32 (42.7) 42 (37.8) Best supportive care 21 (11.3) 0 (0.0) 21 (18.9) Chemotherapy 62 (33.3) 0 (0.0) 62 (55.9) Initial first-line therapy, n (%) SBRT or IAT with or w/o 28 (15.1) 0 (0.0) 28 (25.2) chemotherapy Surgery 75 (40.3) 75 (100.0) 0 (0.0) Lung 105 (56.5) 40 (53.3) 65 (58.6) Peritoneum 81 (43.5) 30 (40.0) 51 (45.9) Type of distant metastasis, n (%) Bone 50 (26.9) 18 (24.0) 32 (28.8) Other 24 (12.9) 12 (16.0) 12 (10.8) At first metastatic occurrence 27 (14.5) 6 (8.0) 21 (18.9) Multiple affected metastatic sites, n (%) Over the disease course 58 (31.2) 19 (25.3) 39 (35.1) Sum of intrahepatic lesions , mm, 103 [47–167] 42 [8–101] 150 [86–191] median (IQR) EEM 18 (9.7) 18 (24.0) 0 (0.0) <25% 95 (51.1) 47 (62.7) 48 (43.3) Hepatic tumor burden , n (%) 25–50% 53 (28.5) 9 (12.0) 44 (39.6) >50% 20 (10.7) 1 (1.3) 19 (17.1) CA 19-9 serum levels , U/mL, median 107 37 [16–261] 137 [32–1862] (IQR) [22–1442] Albumin , g/L, median (IQR) 35 [31–39] 36 [33–40] 34 [29–39] Best supportive care 63 (33.9) 17 (22.7) 46 (41.5) Chemotherapy 95 (51.1) 48 (64.0) 47 (42.3) Subsequent therapy , n (%) SBRT or IAT with or w/o 21 (11.3) 3 (4.0) 18 (16.2) chemotherapy Surgery 7 (3.7) 7 (9.3) 0 IQR, interquartile range; EEM, exclusively extrahepatic metastases; CA 19-9, carbohydrate antigen 19-9; SBRT, stereotactic body radiotherapy; IAT, intra- † ‡ arterial therapy. ,e sum of distant metastases is >100% because patients could have more than one metastatic site. Other sites include the brain, adrenals, spleen, and soft tissues/skin (cutaneous metastases). At time of first metastatic occurrence. metastases is a rare event that has only been described in case Regarding subsequent treatment, it is important to note reports [22]. ,e number of patients who developed brain that this factor is influenced by a multitude of factors in- metastases in our study was very small (n � 3), and the cluding performance status, liver function, and previous treatments. ,us, a selection bias is inherent, and it is not aggregated tumor location “other” did not show a signifi- cantly worse outcome compared with the three main sites. surprising that patients who received merely best supportive care once metastases were detected showed the worst Hepatic tumor burden was the strongest predictor of poor survival in our study. Formerly resected patients with prognosis compared with other treatment groups [29]. exclusively extrahepatic metastases showed the best prog- However, it is important to stress that hepatic tumor burden nosis of 16.3 months, whereas patients with more than 50% maintained its role as the strongest predictor of survival in hepatic tumor infestation had an abysmal mean OS of only multivariate Cox regression even when subsequent treat- 1.2 months (P< 0.001). In previous studies investigating ment was included as a cofactor. ICC, multifocality and tumor size were established as risk Systemic chemotherapy is the mainstay of palliative factors [12, 23]. In other tumor entities, a tumor burden treatment [7]; however, intra-arterial therapies are consid- score based on tumor size and number was also recently ered on an individual basis. While single-institution studies for TACE and SIRT are promising [30, 31], systematic re- proposed for risk stratification for patients with colorectal liver metastasis, demonstrating good prognostic discrimi- views stress the heterogeneity among studies and the vari- natory power in internal and external validation [24]. ance in patient selection and indications [32]. Prospective In addition to hepatic tumor burden, low serum albumin randomized controlled trials comparing intra-arterial was also associated with increased mortality in our study. therapy with systemic treatment are still ongoing (e.g., While the mechanism of protein synthesis and distribution SIRCCA trial, ClinicalTrials.gov Identifier: NCT02807181). is complex, albumin levels have been associated with liver A small number of patients underwent surgical treat- function and patient survival in critically ill patients and ment despite the presence of distant metastases (n � 7). In cancer patients [25–27]. Recently, high serum albumin was most of these cases, liver resection was performed despite the presence of small pulmonary nodules, which were later also found to confer a survival advantage in patients with hilar cholangiocarcinoma [28]. confirmed to be metastases in follow-up imaging (n � 4). Journal of Oncology 5 0 1224364860 Time difference to initial diagnosis (months) Type Lung Bone Peritoneum Other Figure 3: Histogram depicting time differences between initial diagnosis and metastatic occurrence for each site. 1.00 + a cutaneous metastasis (n � 1) respectively. However, in nonmetastasized patients, there has been emerging evidence supporting re-resection for recurrent ICC [33]. Moreover, 0.75 Zhang et al. [34] reported a survival benefit for Asian pa- tients with metastasized hepatolithiasis-associated ICC 0.50 undergoing palliative resection. Yamada et al. [35] published a case report of long-term survival after surgical resection for + + recurrent hepatic and pulmonary metastases of ICC. 0.25 + ,erefore, there might be room for advocating liver resec- tion when the extrahepatic manifestations are limited. Our analysis had several limitations. First and foremost, 0.00 the study was single-centered and conducted in a retro- 0 12 24 36 48 60 spective fashion. ,erefore, a bias due to center-specific Time (months) treatment may be present. Independent external validation Number at risk Lung 105 26 8 62 2 of the results presented in this study is still missing, and Peritoneum 81 18 3 2 0 0 future multicenter approaches may be necessary to validate Bone 50 8 1 1 0 0 Other 24 4 2 0 0 0 our results. Second, the number of investigated patients 012 24 36 48 60 (n � 370) was moderate. However, given the low incidence of Time (months) ICC in Western countries, existing studies investigating patients with ICC have similar or even considerably smaller Strata sample sizes. Due to the long study period, both improved + Lung + Bone imaging technologies and improved treatment options over + Peritoneum + Other time carry a bias that is difficult to control for. Availability Figure 4: Kaplan–Meier curves of OS stratified according to and quality of cross-sectional imaging have improved over metastatic sites. the years, with more stringent follow-ups towards the end of the recruitment period. ,erefore, a selections bias towards Furthermore, liver resection was performed once in the the later years in the recruitment period is present and knowledge of a bone metastasis (n � 1). Two patients un- patients with metastatic disease might have been missed in derwent surgery to excise a peritoneal metastasis (n � 1) and earlier years. Furthermore, patients received different Strata Survival probability Counts Other Bone Peritoneum Lung 6 Journal of Oncology 1.00 + 0.75 0.50 0.25 0.00 0 12 24 36 48 60 Time (months) Number at risk EEM 18 12 5 30 0 <25% 95 28 7 4 0 0 25−50% 53 9 2 22 2 >50% 20 2 0 0 0 0 0 12 24 36 48 60 Time (months) Strata + EEM + 25–50% + <25% + >50% Figure 5: Kaplan–Meier curves of OS stratified according to hepatic tumor burden at the time of first metastatic occurrence (EEM, exclusively extrahepatic metastases). Variable N Hazard ratio P Metastasis type Lung 78 Reference Peritoneum 55 1.22 (0.81, 1.83) 0.35 Bone 36 1.37 (0.86, 2.18) 0.19 Other 17 1.54 (0.86, 2.77) 0.15 Metastatic sites Single 159 Reference Multiple 27 1.52 (0.89, 2.60) 0.13 Hepatic burden EEM 18 Reference <25% 95 2.23 (1.20, 4.14) 0.01 25–50% 53 4.03 (1.90, 8.55) <0.001 >50% 20 9.07 (3.84, 21.46) <0.001 Recurrence No 118 Reference Yes 68 1.27 (0.80, 2.01) 0.31 Subsequent treatment Surgery 7 Reference ∗∗ SBRT or IAT 21 0.63 (0.26, 1.57) 0.33 Cx 95 0.98 (0.42, 2.27) 0.96 BSC 63 2.93 (1.27, 6.76) 0.01 CA 19–9 ≤37 U/mL 67 Reference >37 U/mL 119 1.13 (0.80, 1.59) 0.49 Albumin >35 g/L 89 Reference ≤35 g/L 97 2.37 (1.68, 3.33) <0.001 0.5 1 2 5 10 20 Figure 6: Multivariate Cox hazard regression for survival from the time of first metastatic occurrence (EEM, exclusively extrahepatic metastases; SBRT, stereotactic body radiotherapy; IAT, intra-arterial therapy; Cx, chemotherapy; BSC, best supportive care; CA 19-9, carbohydrate antigen 19-9). In case of simultaneous metastatic spread to two or more different sites at the time of first metastatic ∗∗ occurrence, the more rarely observed metastasis was counted. In combination with or without chemotherapy. chemotherapy regimen over time; the current standard of since then due to adherence to this regimen. Moreover, there gemcitabine and cisplatin was introduced in 2010 following were no stipulated treatment options once metastases were the UK-ABC2 trial [7], and outcomes might have improved detected, but therapeutic concepts were devised in Strata Survival probability Journal of Oncology 7 interdisciplinary tumor boards for each patient on an in- Supplementary Materials dividual basis. Figure S1: Kaplan–Meier curves of OS stratified according to presence of distant metastases at initial diagnosis. Figure S2: 5. Conclusions Kaplan–Meier curves of residual OS at time of recurrence While the presence of distant metastases is frequent in stratified according to presence of distant metastases. patients with ICC and is associated with a poor outcome, (Supplementary Materials) there was no discernible difference in OS between metastatic sites. However, hepatic tumor burden was the life-limiting References risk factor in a majority of patients. ,erefore, in addition to [1] J. L. Petrick, M. Braunlin, M. Laversanne, P. C. Valery, F. Bray, chemotherapy, interdisciplinary approaches including re- and K. A. McGlynn, “International trends in liver cancer section and intra-arterial therapy might be considerations incidence, overall and by histologic subtype, 1978–2007,” on an individual basis to achieve hepatic tumor control even International Journal of Cancer, vol. 139, no. 7, pp. 1534–1545, in the presence of distant metastases if deemed oncologically reasonable. [2] Y. H. Shaib, J. A. Davila, K. McGlynn, and H. B. El-Serag, “Rising incidence of intrahepatic cholangiocarcinoma in the Abbreviations United States: a true increase?” Journal of Hepatology, vol. 40, no. 3, pp. 472–477, 2004. ICC: Intrahepatic cholangiocarcinoma [3] J. D. Yang, B. Kim, S. O. Sanderson et al., “Biliary tract cancers HCC: Hepatocellular carcinoma in olmsted county, minnesota, 1976–2008,” American Journal TACE: Transarterial chemoembolization of Gastroenterology, vol. 107, no. 8, pp. 1256–1262, 2012. [4] T. von Hahn, S. Ciesek, G. Wegener et al., “Epidemiological SIRT: Selective internal radiation therapy trends in incidence and mortality of hepatobiliary cancers in SEER: Surveillance, Epidemiology, and End Results Germany,” Scandinavian Journal of Gastroenterology, vol. 46, OS: Overall survival pp. 1092–1098, 2011. CT: Computed tomography [5] H. Guro, J. W. Kim, Y. Choi, J. Y. Cho, Y.-S. Yoon, and EEM: Exclusively extrahepatic metastases H.-S. Han, “Multidisciplinary management of intrahepatic HR: Hazard ratio. cholangiocarcinoma: current approaches,” Surgical Oncology, vol. 26, no. 2, pp. 146–152, 2017. Data Availability [6] J. Park, M.-H. Kim, K.-p. Kim et al., “Natural history and prognostic factors of advanced cholangiocarcinoma without ,e data used to support the findings of this study are surgery, chemotherapy, or radiotherapy: a large-scale ob- available from the corresponding author upon request. servational study,” Gut and Liver, vol. 3, no. 4, pp. 298–305, [7] J. Valle, H. Wasan, D. H. Palmer et al., “Cisplatin plus Ethical Approval gemcitabine versus gemcitabine for biliary tract cancer,” New England Journal of Medicine, vol. 362, no. 14, pp. 1273–1281, ,e study was approved by the responsible ethics committee (Ethics committee of the Medical Association of Rhineland [8] H. M. Park, S. P. Yun, E. C. Lee et al., “Outcomes for patients Palatinate, Mainz, Germany) for the retrospective analysis of with recurrent intrahepatic cholangiocarcinoma after sur- clinical data (permit number 2018-13618). Additional ex- gery,” Annals of Surgical Oncology, vol. 23, no. 13, aminations were not performed. Patient records and in- pp. 4392–4400, 2016. formation were anonymized and deidentified prior to [9] L. M. Boehm, T. T. Jayakrishnan, J. T. Miura et al., “Com- analysis. parative effectiveness of hepatic artery based therapies for unresectable intrahepatic cholangiocarcinoma,” Journal of Disclosure Surgical Oncology, vol. 111, no. 2, pp. 213–220, 2015. [10] S. M. Weber, D. Ribero, E. M. O’Reilly, N. Kokudo, M. Miyazaki, Preliminary results of this manuscript have been presented and T. M. Pawlik, “Intrahepatic cholangiocarcinoma: expert as electronic poster at ECR 2020 (poster no.: C-01185, consensus statement,” HBP-International Hepato-Pancreato-Bil- iary Association, vol. 17, no. 8, pp. 669–680, 2015. https://doi.org/10.26044/ecr2020/C-01185). [11] A. A. Rahnemai-Azar, A. B. Weisbrod, M. Dillhoff, C. Schmidt, and T. M. Pawlik, “Intrahepatic chol- Conflicts of Interest angiocarcinoma: current management and emerging thera- pies,” Expert Review of Gastroenterology & Hepatology, vol. 11, PRG has received grants and personal fees from Bayer and no. 5, pp. 439–449, 2017. personal fees from Bristol-Myers Squibb, MSD Sharp & [12] Y. Wang, J. Li, Y. Xia et al., “Prognostic nomogram for Dohme, Lilly, SillaJen, SIRTEX, and AstraZeneca. AW has intrahepatic cholangiocarcinoma after partial hepatectomy,” received speaker fees and travel grants from Bayer. DPDS Journal of Clinical Oncology, vol. 31, no. 9, pp. 1188–1195, has received personal fees from Cook. RK has received speaker fees from BTG, Guerbet, and SIRTEX and personal [13] Z.-W. Meng, W. Pan, H.-J. Hong, J.-Z. Chen, and Y.-L. Chen, fees from Boston Scientific, Bristol-Myers Squibb, Guerbet, “Macroscopic types of intrahepatic cholangiocarcinoma and and SIRTEX. None of these companies supported this study, the eighth edition of AJCC/UICC TNM staging system,” and none of the authors report any conflict of interest. Oncotarget, vol. 8, no. 60, pp. 101165–101174, 2017. 8 Journal of Oncology [14] F. Hahn, L. Muller, ¨ F. Stohr ¨ et al., “,e role of sarcopenia in intra-hepatic cholangiocarcinoma: a meta-regression study,” Liver International, vol. 37, no. 7, pp. 1056–1064, 2017. patients with intrahepatic cholangiocarcinoma: prognostic marker or hyped parameter?” Liver International, vol. 39, [33] F. Bartsch, M. Paschold, J. Baumgart, M. Hoppe-Lotichius, S. Heinrich, and H. Lang, “Surgical resection for recurrent no. 7, pp. 1307–1314, 2019. [15] W. Wu, X. He, D. Andayani et al., “Pattern of distant ex- intrahepatic cholangiocarcinoma,” World Journal of Surgery, vol. 43, no. 4, pp. 1105–1116, 2019. trahepatic metastases in primary liver cancer: a SEER based [34] G.-w. Zhang, J.-h. Lin, J.-p. Qian, and J. Zhou, “Identification study,” Journal of Cancer, vol. 8, no. 12, pp. 2312–2318, 2017. of prognostic factors and the impact of palliative resection on [16] R Core Team, R: A Language and Environment for Statistical survival of patients with stage IV hepatolithiasis-associated Computing, R Core Team, Vienna, Austria, 2018. intrahepatic cholangiocarcinoma,” Journal of Surgical On- [17] G. Frega, I. Garajova, ´ A. Palloni et al., “Brain metastases from cology, vol. 109, no. 5, pp. 494–499, 2014. biliary tract cancer: a monocentric retrospective analysis of [35] M. Yamada, A. Arimoto, Y. Toyoda et al., “Long-term survival 450 patients,” Oncology, vol. 94, no. 1, pp. 7–11, 2018. after surgical resection for recurrent hepatic and pulmonary [18] M. Al-Sarraf, T. S. Go, K. Kithier, and V. K. Vaitkevicius, metastases of intrahepatic cholangiocarcinoma: a case report,” “Primary liver cancer: A review of the clinical features, blood Surgical Case Reports, vol. 5, no. 1, p. 135, 2019. groups, serum enzymes, therapy, and survival of 65 cases,” Cancer, vol. 33, no. 2, pp. 574–582, 1974. [19] Y.-T. M. Lee and D. A. Geer, “Primary liver cancer: pattern of metastasis,” Journal of Surgical Oncology, vol. 36, no. 1, pp. 26–31, 1987. [20] T. Tobe, “Primary liver cancer in Japan. ,e liver cancer study group of Japan,” Cancer, vol. 54, no. 8, pp. 1747–1755, 1984. [21] H. Witzigmann, H. Lang, and H. Lauer, “Guidelines for palliative surgery of cholangiocarcinoma,” HBP-International Hepato-Pancreato-Biliary Association, vol. 10, no. 3, pp. 154–160, 2008. [22] Y. Nagayoshi, K. Yamamoto, S. Hashimoto et al., “An autopsy case of lepidic pulmonary metastasis from chol- angiocarcinoma,” Internal Medicine, vol. 55, no. 19, pp. 2849–2853, 2016. [23] W. Jiang, Z.-C. Zeng, Z. Y. Tang et al., “A prognostic scoring system based on clinical features of intrahepatic chol- angiocarcinoma: the fudan score,” Annals of Oncology, vol. 22, no. 7, pp. 1644–1652, 2011. [24] K. Sasaki, D. Morioka, S. Conci et al., “,e tumor burden score,” Annals of Surgery, vol. 267, no. 1, pp. 132–141, 2018. [25] M. A. Rothschild, M. Oratz, and S. S. Schreiber, “Albumin synthesis,” New England Journal of Medicine, vol. 286, no. 14, pp. 748–757, 1972. [26] J. P. Nicholson, M. R. Wolmarans, and G. R. Park, “,e role of albumin in critical illness,” British Journal of Anaesthesia, vol. 85, no. 4, pp. 599–610, 2000. [27] P. O’Gorman, D. C. McMillan, and C. S. McArdle, “Longi- tudinal study of weight, appetite, performance status, and inflammation in advanced gastrointestinal cancer,” Nutrition and Cancer, vol. 35, no. 2, pp. 127–129, 1999. [28] A. Waghray, A. Sobotka, C. R. Marrero, B. Estfan, F. Aucejo, and K. Narayanan Menon, “Serum albumin predicts survival in patients with hilar cholangiocarcinoma,” Gastroenterology Report, vol. 5, pp. 62–66, 2017. [29] R. Dhanasekaran, A. W. Hemming, I. Zendejas et al., “Treatment outcomes and prognostic factors of intrahepatic cholangiocarcinoma,” Oncology Reports, vol. 29, no. 4, pp. 1259–1267, 2013. [30] N. J. Gusani, F. K. Balaa, J. L. Steel et al., “Treatment of unresectable cholangiocarcinoma with gemcitabine-based transcatheter arterial chemoembolization (TACE): a single- institution experience,” Journal of Gastrointestinal Surgery, vol. 12, no. 1, pp. 129–137, 2008. [31] T. M. Shaker, C. Chung, M. K. Varma et al., “Is there a role for ytrrium-90 in the treatment of unresectable and metastatic intrahepatic cholangiocarcinoma?” 6e American Journal of Surgery, vol. 215, no. 3, pp. 467–470, 2018. [32] A. Cucchetti, A. Cappelli, C. Mosconi et al., “Improving patient selection for selective internal radiation therapy of http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Oncology Hindawi Publishing Corporation

Distant Metastases in Patients with Intrahepatic Cholangiocarcinoma: Does Location Matter? A Retrospective Analysis of 370 Patients

Loading next page...
 
/lp/hindawi-publishing-corporation/distant-metastases-in-patients-with-intrahepatic-cholangiocarcinoma-hPymkGIwUh

References (41)

Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2020 Felix Hahn et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
1687-8450
eISSN
1687-8469
DOI
10.1155/2020/7195373
Publisher site
See Article on Publisher Site

Abstract

Hindawi Journal of Oncology Volume 2020, Article ID 7195373, 8 pages https://doi.org/10.1155/2020/7195373 Research Article Distant Metastases in Patients with Intrahepatic Cholangiocarcinoma: Does Location Matter? A Retrospective Analysis of 370 Patients 1 1 1 2 Felix Hahn , Lukas Mu¨ller, Aline Ma¨hringer-Kunz, Yasemin Tanyildizi, 3 1 4 4,5 Daniel Pinto dos Santos, Christoph Du¨ber, Peter R. Galle, Arndt Weinmann, and Roman Kloeckner Department of Diagnostic and Interventional Radiology, University Medical Center of the Johannes Gutenberg-University Mainz, Mainz, Germany Department of Neuroradiology, University Medical Center of the Johannes Gutenberg-University Mainz, Mainz, Germany Department of Radiology, University Hospital of Cologne, Cologne, Germany Department of Internal Medicine, University Medical Center of the Johannes Gutenberg-University Mainz, Mainz, Germany Clinical Registry Unit (CRU), University Medical Center of the Johannes Gutenberg-University Mainz, Mainz, Germany Correspondence should be addressed to Felix Hahn; felix.hahn@unimedizin-mainz.de Received 18 May 2020; Revised 13 August 2020; Accepted 28 September 2020; Published 10 October 2020 Academic Editor: Dali Zheng Copyright © 2020 Felix Hahn et al. ,is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background. Intrahepatic cholangiocarcinoma (ICC) is an aggressive tumor entity, and distant metastases are common. However, studies investigating patterns and clinical relevance of distant metastases are rare. ,erefore, we aimed to analyze occurrence, location, and prognostic impact of distant metastases on overall survival (OS). Methods. Between 1997 and 2018, 417 patients with ICC were treated at our tertiary care center. Distant metastases and intrahepatic tumor burden were retrospectively evaluated in a longitudinal approach using volumetric assessment of cross-sectional imaging studies and all available medical/histopathological reports. Results. Finally, 370 patients with histopathologically confirmed ICC were included. Of these, 186 showed distant metastases, either initially (n � 59) or during follow-up (n � 127). ,e most common metastatic sites were the lung (n � 105), peritoneum (n � 81), and bone (n � 50). After detection of lung metastases, the residual median OS was 5.3 months; followed by peritoneal metastases, 4.5 months, and bone metastases, 4.4 months (P � 0.17). At the time of first metastatic occurrence, residual OS according to intrahepatic tumor burden of <25%, 25–50%, and >50% was 6.5 months, 4.9 months, and 1.2 months, re- spectively (P< 0.001). In multivariate hazard regression, hepatic tumor burden, liver function, and subsequent treatment were significant predictors of survival. Conclusions. During the disease course, every second patient developed extrahepatic metastases. While the presence of distant metastases was associated with poor patient outcomes, there was no significant difference between metastatic sites. However, hepatic tumor burden was the life-limiting risk factor in a majority of patients at the time of distant metastatic disease. Affected patients are often asymptomatic in the early 1. Introduction stages; thus at diagnosis, the tumor is often already at an Intrahepatic cholangiocarcinoma (ICC) is the second most advanced stage [5]. Nonresectable patients have repeatedly common primary liver malignancy after hepatocellular been identified as having a poorer prognosis than patients carcinoma (HCC). Incidence in Western countries is esti- undergoing surgery [6, 7]. However, even if resection is mated to be approximately 0.4–2.0/100,000 and has con- possible, tumor recurrence has been reported in about 65% siderably increased since the last three decades [1–4]. of the patients [8]. 2 Journal of Oncology Regarding systemic chemotherapy, the combination of gemcitabine and cisplatin has been widely administered as first-line chemotherapy since the publication of the multi- center UK-ABC 02 study in 2010 [7]. Even though systemic chemotherapy is the mainstay of treatment, intra-arterial therapies such as transarterial chemoembolization (TACE) or selective internal radiation therapy (SIRT) have also been established since the last decade as treatment options for selected patients [9]. However, prognosis remains poor [10]. Figure 1: Exemplary contrast-enhanced axial CTslices in the portal venous phase depicting the volumetric measurement of tumor-free As resection is the only curative treatment option, the liver tissue for one patient using Aquarius iNtuition software. presence of distant metastases impedes curative-intent re- sections [11]. Distant metastases are incorporated into the prognostic Wang nomogram [12]; also, the discriminator th 25–50%, >50%). Patients after resection without hepatic between stage III and stage IV in the current 8 edition of recurrence but distant metastases were categorized as “ex- the UICC staging system is the presence of distant metas- clusively extrahepatic metastases” (EEM). tases [13]. In a recent study performed by our group, distant Occurrence and distribution of distant metastases metastases were independent predictors of poor survival as were evaluated using the statistical software package R well [14]. 3.5.1 [16]. Survival analyses were performed using the Based on the Surveillance, Epidemiology, and End Re- “survival” and “survminer” (https://cran.r-project.org/ sults (SEER) database, Wu et al. [15] investigated patterns of package�survival, https://cran.r-project.org/ distant extrahepatic metastases in primary liver cancer, both package�survminer, accessed on 31.12.2019) packages. HCC and ICC. However, the authors limited their inves- In rare cases of missing laboratory values, these were tigation to patients with distant metastases at the time of imputed using the “mice” package (https://cran.r-project. diagnosis only. Moreover, survival analyses were presented org/package�mice, accessed on 31.12.2019). Log-rank for pooled HCC and ICC patients, with ICC patients ac- tests and Kaplan–Meier curves were used for survival counting for only 11% of the study population. analysis between strata. Univariate and multivariate Cox To the best of our knowledge, no study has focused on proportional hazards regression models were fitted to the longitudinal investigation of metastatic occurrence and determine the influence of predictors. As this analysis has the impact of different sites of metastatic spread for patients exploratory intention, P-values should be interpreted in a with ICC. ,erefore, the aim of this study was to analyze descriptive manner. P values< 0.05 were considered occurrence, location, and prognostic impact of distant statistically significant. metastases on overall survival (OS) during the course of disease. 3. Results 2. Materials and Methods 3.1. Occurrence and Location of Distant Metastases. Of the Between January 1997 and January 2018, 417 patients with 417 patients, 47 had to be excluded for reasons as de- histopathologically confirmed ICC were treated at our scribed in the STROBE flowchart (Figure 2); the tertiary care center. ,ese patients were retrospectively remaining 370 patients were included in this study. A identified from a dedicated, prospectively populated clinical total of 200/370 patients underwent primary surgical database. Data from follow-up visits were extracted from the resection, while 170/370 patients were considered non- hospital and radiology information systems. Death dates resectable by an interdisciplinary board due to their were queried at the appropriate resident’s registration of- general state of health or advanced tumor stage asserted fices. Follow-up was ended on December 31, 2018. In case of by imaging studies or surgical exploration. Of the 200 loss to follow-up, patients were censored at the date of last patients undergoing resection, 7 had distant metastases at contact. ,e study was approved by the responsible ethics initial diagnosis, while 68 developed distant metastases committee for the retrospective analysis of clinical data during follow-up. Of the 170 nonresectable patients, 52 (Permit No. 2018–13618). initially presented with distant metastases and 59 de- We evaluated contrast-enhanced computed tomography veloped distant metastases during the course of disease. (CT) or magnetic resonance imaging at diagnosis and during Further baseline characteristics of the patients at the time the patients’ course of disease to determine the size and of the first development of distant metastases are depicted number of intrahepatic lesions as well as the presence of in Table 1. distant metastases in the lung, the peritoneum, bones, other ,e most common metastatic site was the lung (n � 105, organs, and soft tissues other than lymph nodes. At the time 28%), followed by the peritoneum (n � 81, 22%) and bones of first development of distant metastases, hepatic tumor (n � 50, 14%). Other metastatic sites included the adrenals burden as percentage of total liver parenchyma was mea- (n � 8, 2%), brain (n � 3, 1%), spleen (n � 5, 1%), and soft sured by volumetric assessment using dedicated third-party tissues/skin (cutaneous metastases) (n � 8, 2%); these sites software (Aquarius iNtuition ; TeraRecon, Foster City, CA, are henceforth referred to as “other.” ,e time distribution USA; Figure 1) and categorized into three groups (<25%, of metastatic occurrence for each site is depicted in Figure 3. Journal of Oncology 3 recurrence (Figure S2, median residual OS 6.4 months vs. Patients with 20.6 months, P � 0.01). histopathologically In multivariate Cox hazard regression, hepatic tumor confirmed ICC (n = 417) burden, liver function, and subsequent treatment were significant predictors of survival (Figure 6). While the Cross-sectional imaging categories of “bone” and “other” metastases showed an at initial diagnosis missing (n = 36) increased hazard, the effect was not significant. Hepatic tumor burden was the factor with the highest hazard ratios (HRs) in multivariate analysis. Lost to follow-up early after discharge (n = 6) 4. Discussion Cross-sectional imaging In our cohort, the number of patients with extrahepatic at time of metastatic metastases was high with every second patient developing disease missing (n = 5) extrahepatic metastases during the course of disease. Distant metastases were associated with poor outcome, irrespective Finally included of the metastatic site. ,e rarer metastatic locations showed (n = 370) an increased hazard; however, the effect was not significant. In contrast, the amount of hepatic tumor burden and the patient’s ability to tolerate further treatment were the strongest predictors of survival in multivariate Cox hazard Metastatic disease No metastases in regression analysis. (n = 186) follow-up (n = 184) Literature on the investigated topic is scarce, particularly Figure 2: STROBE flowchart showing the number and the reasons regarding the longitudinal investigation of metastatic oc- for dropout. currence and different sites of metastatic spread. ,e SEER- based study by Wu et al. [15] investigated patterns of distant extrahepatic metastases at initial diagnosis in both HCC and 3.2. Survival of Patients with Distant Metastases. When in- ICC. However, they only present pooled survival analyses for vestigating survival from the time of detection of metastatic both cancer entities. Moreover, while the authors claim in spread until death or loss to follow-up, all metastatic sites their conclusion that there were profound differences in risk showed similar survival curves without any statistically of mortality among distant extrahepatic metastatic sites, the significant difference. After detection of lung metastases, the presented confidence intervals of HRs overlap for lung, median OS was 5.3 months; after peritoneal metastases, it bone, and distant lymph node metastases, with only the HR was 4.5 months; and after bone metastases, it was 4.4 of brain metastases significantly differing from the other months. Pooled patients with metastases other than the three metastatic sites. aforementioned locations showed an OS of 4.5 months In both the study by Wu et al. [15] and the current study, (Figure 4). Pairwise comparisons between survival times lung metastases were the most common extrahepatic manifestation. However, in their study, the second most using the log-rank test resulted in P values ranging from 0.29 to 0.75, with the lowest P value of 0.29 between patients with common metastases were bone metastases. It is unclear lung metastases and those with bone metastases. whether peritoneal metastases, the second most common However, when stratifying patients according to hepatic metastases in our study, were investigated. Frega et al. [17] tumor burden at the time of first metastatic occurrence, investigated the occurrence of brain metastases from biliary survival was significantly different. ,e residual survival of tract cancer in 450 patients and found an incidence of about formerly resected patients with exclusively extrahepatic 1.4%, with a median OS from detection of brain metastases metastases was 16.3 months compared with 6.5 months, 4.9 of 4 months. ,is corresponds with the subset of patients months, and 1.2 months for patients with <25%, 25–50%, developing brain metastases in our cohort. Older, autopsy- and >50% hepatic tumor burden, respectively (Figure 5). based studies featured only small patient numbers not ex- Pairwise comparisons among survival times resulted in P ceeding 50 patients, and the share of patients affected by values< 0.01 for all combinations except comparing the distant metastases differs considerably among studies. Lung “<25%” with the “25–50%” group, where the P value was metastases have been reported in 20–40%, and bone me- 0.12. tastases in 10–55% of patients [18–20]. ,e results presented Kaplan–Meier curves of patients with metastatic in our study fall within these broad ranges. disease at initial diagnosis compared with patients Even though the survival of patients with ICC was without metastatic spread at initial diagnosis can be generally poor once metastases were detected, there was no found in Supplementary Materials (Figure S1, median OS discernible influence of metastatic sites on survival in our 4.2 months vs. 17.7 months, P< 0.001). Moreover, sur- cohort. ,is corresponds with clinical routine in that liver vival analysis was performed comparing patients with failure, and especially, biliary complications can pose life- metastatic disease at the time of tumor recurrence with threatening situations for patients with advanced ICC [21], patients without metastatic disease at the time of whereas, for example, respiratory failure due to lung 4 Journal of Oncology Table 1: Baseline characteristics of patients with distant metastases. Initially resected Initially nonresectable All (n � 186) (n � 75) (n � 111) Age, years, median (IQR) 63.8 (56–72) 61.3 (55–68) 67.0 (57–74) Male 112 (60.2) 43 (57.3) 69 (62.2) Sex, n (%) Female 74 (39.8) 32 (42.7) 42 (37.8) Best supportive care 21 (11.3) 0 (0.0) 21 (18.9) Chemotherapy 62 (33.3) 0 (0.0) 62 (55.9) Initial first-line therapy, n (%) SBRT or IAT with or w/o 28 (15.1) 0 (0.0) 28 (25.2) chemotherapy Surgery 75 (40.3) 75 (100.0) 0 (0.0) Lung 105 (56.5) 40 (53.3) 65 (58.6) Peritoneum 81 (43.5) 30 (40.0) 51 (45.9) Type of distant metastasis, n (%) Bone 50 (26.9) 18 (24.0) 32 (28.8) Other 24 (12.9) 12 (16.0) 12 (10.8) At first metastatic occurrence 27 (14.5) 6 (8.0) 21 (18.9) Multiple affected metastatic sites, n (%) Over the disease course 58 (31.2) 19 (25.3) 39 (35.1) Sum of intrahepatic lesions , mm, 103 [47–167] 42 [8–101] 150 [86–191] median (IQR) EEM 18 (9.7) 18 (24.0) 0 (0.0) <25% 95 (51.1) 47 (62.7) 48 (43.3) Hepatic tumor burden , n (%) 25–50% 53 (28.5) 9 (12.0) 44 (39.6) >50% 20 (10.7) 1 (1.3) 19 (17.1) CA 19-9 serum levels , U/mL, median 107 37 [16–261] 137 [32–1862] (IQR) [22–1442] Albumin , g/L, median (IQR) 35 [31–39] 36 [33–40] 34 [29–39] Best supportive care 63 (33.9) 17 (22.7) 46 (41.5) Chemotherapy 95 (51.1) 48 (64.0) 47 (42.3) Subsequent therapy , n (%) SBRT or IAT with or w/o 21 (11.3) 3 (4.0) 18 (16.2) chemotherapy Surgery 7 (3.7) 7 (9.3) 0 IQR, interquartile range; EEM, exclusively extrahepatic metastases; CA 19-9, carbohydrate antigen 19-9; SBRT, stereotactic body radiotherapy; IAT, intra- † ‡ arterial therapy. ,e sum of distant metastases is >100% because patients could have more than one metastatic site. Other sites include the brain, adrenals, spleen, and soft tissues/skin (cutaneous metastases). At time of first metastatic occurrence. metastases is a rare event that has only been described in case Regarding subsequent treatment, it is important to note reports [22]. ,e number of patients who developed brain that this factor is influenced by a multitude of factors in- metastases in our study was very small (n � 3), and the cluding performance status, liver function, and previous treatments. ,us, a selection bias is inherent, and it is not aggregated tumor location “other” did not show a signifi- cantly worse outcome compared with the three main sites. surprising that patients who received merely best supportive care once metastases were detected showed the worst Hepatic tumor burden was the strongest predictor of poor survival in our study. Formerly resected patients with prognosis compared with other treatment groups [29]. exclusively extrahepatic metastases showed the best prog- However, it is important to stress that hepatic tumor burden nosis of 16.3 months, whereas patients with more than 50% maintained its role as the strongest predictor of survival in hepatic tumor infestation had an abysmal mean OS of only multivariate Cox regression even when subsequent treat- 1.2 months (P< 0.001). In previous studies investigating ment was included as a cofactor. ICC, multifocality and tumor size were established as risk Systemic chemotherapy is the mainstay of palliative factors [12, 23]. In other tumor entities, a tumor burden treatment [7]; however, intra-arterial therapies are consid- score based on tumor size and number was also recently ered on an individual basis. While single-institution studies for TACE and SIRT are promising [30, 31], systematic re- proposed for risk stratification for patients with colorectal liver metastasis, demonstrating good prognostic discrimi- views stress the heterogeneity among studies and the vari- natory power in internal and external validation [24]. ance in patient selection and indications [32]. Prospective In addition to hepatic tumor burden, low serum albumin randomized controlled trials comparing intra-arterial was also associated with increased mortality in our study. therapy with systemic treatment are still ongoing (e.g., While the mechanism of protein synthesis and distribution SIRCCA trial, ClinicalTrials.gov Identifier: NCT02807181). is complex, albumin levels have been associated with liver A small number of patients underwent surgical treat- function and patient survival in critically ill patients and ment despite the presence of distant metastases (n � 7). In cancer patients [25–27]. Recently, high serum albumin was most of these cases, liver resection was performed despite the presence of small pulmonary nodules, which were later also found to confer a survival advantage in patients with hilar cholangiocarcinoma [28]. confirmed to be metastases in follow-up imaging (n � 4). Journal of Oncology 5 0 1224364860 Time difference to initial diagnosis (months) Type Lung Bone Peritoneum Other Figure 3: Histogram depicting time differences between initial diagnosis and metastatic occurrence for each site. 1.00 + a cutaneous metastasis (n � 1) respectively. However, in nonmetastasized patients, there has been emerging evidence supporting re-resection for recurrent ICC [33]. Moreover, 0.75 Zhang et al. [34] reported a survival benefit for Asian pa- tients with metastasized hepatolithiasis-associated ICC 0.50 undergoing palliative resection. Yamada et al. [35] published a case report of long-term survival after surgical resection for + + recurrent hepatic and pulmonary metastases of ICC. 0.25 + ,erefore, there might be room for advocating liver resec- tion when the extrahepatic manifestations are limited. Our analysis had several limitations. First and foremost, 0.00 the study was single-centered and conducted in a retro- 0 12 24 36 48 60 spective fashion. ,erefore, a bias due to center-specific Time (months) treatment may be present. Independent external validation Number at risk Lung 105 26 8 62 2 of the results presented in this study is still missing, and Peritoneum 81 18 3 2 0 0 future multicenter approaches may be necessary to validate Bone 50 8 1 1 0 0 Other 24 4 2 0 0 0 our results. Second, the number of investigated patients 012 24 36 48 60 (n � 370) was moderate. However, given the low incidence of Time (months) ICC in Western countries, existing studies investigating patients with ICC have similar or even considerably smaller Strata sample sizes. Due to the long study period, both improved + Lung + Bone imaging technologies and improved treatment options over + Peritoneum + Other time carry a bias that is difficult to control for. Availability Figure 4: Kaplan–Meier curves of OS stratified according to and quality of cross-sectional imaging have improved over metastatic sites. the years, with more stringent follow-ups towards the end of the recruitment period. ,erefore, a selections bias towards Furthermore, liver resection was performed once in the the later years in the recruitment period is present and knowledge of a bone metastasis (n � 1). Two patients un- patients with metastatic disease might have been missed in derwent surgery to excise a peritoneal metastasis (n � 1) and earlier years. Furthermore, patients received different Strata Survival probability Counts Other Bone Peritoneum Lung 6 Journal of Oncology 1.00 + 0.75 0.50 0.25 0.00 0 12 24 36 48 60 Time (months) Number at risk EEM 18 12 5 30 0 <25% 95 28 7 4 0 0 25−50% 53 9 2 22 2 >50% 20 2 0 0 0 0 0 12 24 36 48 60 Time (months) Strata + EEM + 25–50% + <25% + >50% Figure 5: Kaplan–Meier curves of OS stratified according to hepatic tumor burden at the time of first metastatic occurrence (EEM, exclusively extrahepatic metastases). Variable N Hazard ratio P Metastasis type Lung 78 Reference Peritoneum 55 1.22 (0.81, 1.83) 0.35 Bone 36 1.37 (0.86, 2.18) 0.19 Other 17 1.54 (0.86, 2.77) 0.15 Metastatic sites Single 159 Reference Multiple 27 1.52 (0.89, 2.60) 0.13 Hepatic burden EEM 18 Reference <25% 95 2.23 (1.20, 4.14) 0.01 25–50% 53 4.03 (1.90, 8.55) <0.001 >50% 20 9.07 (3.84, 21.46) <0.001 Recurrence No 118 Reference Yes 68 1.27 (0.80, 2.01) 0.31 Subsequent treatment Surgery 7 Reference ∗∗ SBRT or IAT 21 0.63 (0.26, 1.57) 0.33 Cx 95 0.98 (0.42, 2.27) 0.96 BSC 63 2.93 (1.27, 6.76) 0.01 CA 19–9 ≤37 U/mL 67 Reference >37 U/mL 119 1.13 (0.80, 1.59) 0.49 Albumin >35 g/L 89 Reference ≤35 g/L 97 2.37 (1.68, 3.33) <0.001 0.5 1 2 5 10 20 Figure 6: Multivariate Cox hazard regression for survival from the time of first metastatic occurrence (EEM, exclusively extrahepatic metastases; SBRT, stereotactic body radiotherapy; IAT, intra-arterial therapy; Cx, chemotherapy; BSC, best supportive care; CA 19-9, carbohydrate antigen 19-9). In case of simultaneous metastatic spread to two or more different sites at the time of first metastatic ∗∗ occurrence, the more rarely observed metastasis was counted. In combination with or without chemotherapy. chemotherapy regimen over time; the current standard of since then due to adherence to this regimen. Moreover, there gemcitabine and cisplatin was introduced in 2010 following were no stipulated treatment options once metastases were the UK-ABC2 trial [7], and outcomes might have improved detected, but therapeutic concepts were devised in Strata Survival probability Journal of Oncology 7 interdisciplinary tumor boards for each patient on an in- Supplementary Materials dividual basis. Figure S1: Kaplan–Meier curves of OS stratified according to presence of distant metastases at initial diagnosis. Figure S2: 5. Conclusions Kaplan–Meier curves of residual OS at time of recurrence While the presence of distant metastases is frequent in stratified according to presence of distant metastases. patients with ICC and is associated with a poor outcome, (Supplementary Materials) there was no discernible difference in OS between metastatic sites. However, hepatic tumor burden was the life-limiting References risk factor in a majority of patients. ,erefore, in addition to [1] J. L. Petrick, M. Braunlin, M. Laversanne, P. C. Valery, F. Bray, chemotherapy, interdisciplinary approaches including re- and K. A. McGlynn, “International trends in liver cancer section and intra-arterial therapy might be considerations incidence, overall and by histologic subtype, 1978–2007,” on an individual basis to achieve hepatic tumor control even International Journal of Cancer, vol. 139, no. 7, pp. 1534–1545, in the presence of distant metastases if deemed oncologically reasonable. [2] Y. H. Shaib, J. A. Davila, K. McGlynn, and H. B. El-Serag, “Rising incidence of intrahepatic cholangiocarcinoma in the Abbreviations United States: a true increase?” Journal of Hepatology, vol. 40, no. 3, pp. 472–477, 2004. ICC: Intrahepatic cholangiocarcinoma [3] J. D. Yang, B. Kim, S. O. Sanderson et al., “Biliary tract cancers HCC: Hepatocellular carcinoma in olmsted county, minnesota, 1976–2008,” American Journal TACE: Transarterial chemoembolization of Gastroenterology, vol. 107, no. 8, pp. 1256–1262, 2012. [4] T. von Hahn, S. Ciesek, G. Wegener et al., “Epidemiological SIRT: Selective internal radiation therapy trends in incidence and mortality of hepatobiliary cancers in SEER: Surveillance, Epidemiology, and End Results Germany,” Scandinavian Journal of Gastroenterology, vol. 46, OS: Overall survival pp. 1092–1098, 2011. CT: Computed tomography [5] H. Guro, J. W. Kim, Y. Choi, J. Y. Cho, Y.-S. Yoon, and EEM: Exclusively extrahepatic metastases H.-S. Han, “Multidisciplinary management of intrahepatic HR: Hazard ratio. cholangiocarcinoma: current approaches,” Surgical Oncology, vol. 26, no. 2, pp. 146–152, 2017. Data Availability [6] J. Park, M.-H. Kim, K.-p. Kim et al., “Natural history and prognostic factors of advanced cholangiocarcinoma without ,e data used to support the findings of this study are surgery, chemotherapy, or radiotherapy: a large-scale ob- available from the corresponding author upon request. servational study,” Gut and Liver, vol. 3, no. 4, pp. 298–305, [7] J. Valle, H. Wasan, D. H. Palmer et al., “Cisplatin plus Ethical Approval gemcitabine versus gemcitabine for biliary tract cancer,” New England Journal of Medicine, vol. 362, no. 14, pp. 1273–1281, ,e study was approved by the responsible ethics committee (Ethics committee of the Medical Association of Rhineland [8] H. M. Park, S. P. Yun, E. C. Lee et al., “Outcomes for patients Palatinate, Mainz, Germany) for the retrospective analysis of with recurrent intrahepatic cholangiocarcinoma after sur- clinical data (permit number 2018-13618). Additional ex- gery,” Annals of Surgical Oncology, vol. 23, no. 13, aminations were not performed. Patient records and in- pp. 4392–4400, 2016. formation were anonymized and deidentified prior to [9] L. M. Boehm, T. T. Jayakrishnan, J. T. Miura et al., “Com- analysis. parative effectiveness of hepatic artery based therapies for unresectable intrahepatic cholangiocarcinoma,” Journal of Disclosure Surgical Oncology, vol. 111, no. 2, pp. 213–220, 2015. [10] S. M. Weber, D. Ribero, E. M. O’Reilly, N. Kokudo, M. Miyazaki, Preliminary results of this manuscript have been presented and T. M. Pawlik, “Intrahepatic cholangiocarcinoma: expert as electronic poster at ECR 2020 (poster no.: C-01185, consensus statement,” HBP-International Hepato-Pancreato-Bil- iary Association, vol. 17, no. 8, pp. 669–680, 2015. https://doi.org/10.26044/ecr2020/C-01185). [11] A. A. Rahnemai-Azar, A. B. Weisbrod, M. Dillhoff, C. Schmidt, and T. M. Pawlik, “Intrahepatic chol- Conflicts of Interest angiocarcinoma: current management and emerging thera- pies,” Expert Review of Gastroenterology & Hepatology, vol. 11, PRG has received grants and personal fees from Bayer and no. 5, pp. 439–449, 2017. personal fees from Bristol-Myers Squibb, MSD Sharp & [12] Y. Wang, J. Li, Y. Xia et al., “Prognostic nomogram for Dohme, Lilly, SillaJen, SIRTEX, and AstraZeneca. AW has intrahepatic cholangiocarcinoma after partial hepatectomy,” received speaker fees and travel grants from Bayer. DPDS Journal of Clinical Oncology, vol. 31, no. 9, pp. 1188–1195, has received personal fees from Cook. RK has received speaker fees from BTG, Guerbet, and SIRTEX and personal [13] Z.-W. Meng, W. Pan, H.-J. Hong, J.-Z. Chen, and Y.-L. Chen, fees from Boston Scientific, Bristol-Myers Squibb, Guerbet, “Macroscopic types of intrahepatic cholangiocarcinoma and and SIRTEX. None of these companies supported this study, the eighth edition of AJCC/UICC TNM staging system,” and none of the authors report any conflict of interest. Oncotarget, vol. 8, no. 60, pp. 101165–101174, 2017. 8 Journal of Oncology [14] F. Hahn, L. Muller, ¨ F. Stohr ¨ et al., “,e role of sarcopenia in intra-hepatic cholangiocarcinoma: a meta-regression study,” Liver International, vol. 37, no. 7, pp. 1056–1064, 2017. patients with intrahepatic cholangiocarcinoma: prognostic marker or hyped parameter?” Liver International, vol. 39, [33] F. Bartsch, M. Paschold, J. Baumgart, M. Hoppe-Lotichius, S. Heinrich, and H. Lang, “Surgical resection for recurrent no. 7, pp. 1307–1314, 2019. [15] W. Wu, X. He, D. Andayani et al., “Pattern of distant ex- intrahepatic cholangiocarcinoma,” World Journal of Surgery, vol. 43, no. 4, pp. 1105–1116, 2019. trahepatic metastases in primary liver cancer: a SEER based [34] G.-w. Zhang, J.-h. Lin, J.-p. Qian, and J. Zhou, “Identification study,” Journal of Cancer, vol. 8, no. 12, pp. 2312–2318, 2017. of prognostic factors and the impact of palliative resection on [16] R Core Team, R: A Language and Environment for Statistical survival of patients with stage IV hepatolithiasis-associated Computing, R Core Team, Vienna, Austria, 2018. intrahepatic cholangiocarcinoma,” Journal of Surgical On- [17] G. Frega, I. Garajova, ´ A. Palloni et al., “Brain metastases from cology, vol. 109, no. 5, pp. 494–499, 2014. biliary tract cancer: a monocentric retrospective analysis of [35] M. Yamada, A. Arimoto, Y. Toyoda et al., “Long-term survival 450 patients,” Oncology, vol. 94, no. 1, pp. 7–11, 2018. after surgical resection for recurrent hepatic and pulmonary [18] M. Al-Sarraf, T. S. Go, K. Kithier, and V. K. Vaitkevicius, metastases of intrahepatic cholangiocarcinoma: a case report,” “Primary liver cancer: A review of the clinical features, blood Surgical Case Reports, vol. 5, no. 1, p. 135, 2019. groups, serum enzymes, therapy, and survival of 65 cases,” Cancer, vol. 33, no. 2, pp. 574–582, 1974. [19] Y.-T. M. Lee and D. A. Geer, “Primary liver cancer: pattern of metastasis,” Journal of Surgical Oncology, vol. 36, no. 1, pp. 26–31, 1987. [20] T. Tobe, “Primary liver cancer in Japan. ,e liver cancer study group of Japan,” Cancer, vol. 54, no. 8, pp. 1747–1755, 1984. [21] H. Witzigmann, H. Lang, and H. Lauer, “Guidelines for palliative surgery of cholangiocarcinoma,” HBP-International Hepato-Pancreato-Biliary Association, vol. 10, no. 3, pp. 154–160, 2008. [22] Y. Nagayoshi, K. Yamamoto, S. Hashimoto et al., “An autopsy case of lepidic pulmonary metastasis from chol- angiocarcinoma,” Internal Medicine, vol. 55, no. 19, pp. 2849–2853, 2016. [23] W. Jiang, Z.-C. Zeng, Z. Y. Tang et al., “A prognostic scoring system based on clinical features of intrahepatic chol- angiocarcinoma: the fudan score,” Annals of Oncology, vol. 22, no. 7, pp. 1644–1652, 2011. [24] K. Sasaki, D. Morioka, S. Conci et al., “,e tumor burden score,” Annals of Surgery, vol. 267, no. 1, pp. 132–141, 2018. [25] M. A. Rothschild, M. Oratz, and S. S. Schreiber, “Albumin synthesis,” New England Journal of Medicine, vol. 286, no. 14, pp. 748–757, 1972. [26] J. P. Nicholson, M. R. Wolmarans, and G. R. Park, “,e role of albumin in critical illness,” British Journal of Anaesthesia, vol. 85, no. 4, pp. 599–610, 2000. [27] P. O’Gorman, D. C. McMillan, and C. S. McArdle, “Longi- tudinal study of weight, appetite, performance status, and inflammation in advanced gastrointestinal cancer,” Nutrition and Cancer, vol. 35, no. 2, pp. 127–129, 1999. [28] A. Waghray, A. Sobotka, C. R. Marrero, B. Estfan, F. Aucejo, and K. Narayanan Menon, “Serum albumin predicts survival in patients with hilar cholangiocarcinoma,” Gastroenterology Report, vol. 5, pp. 62–66, 2017. [29] R. Dhanasekaran, A. W. Hemming, I. Zendejas et al., “Treatment outcomes and prognostic factors of intrahepatic cholangiocarcinoma,” Oncology Reports, vol. 29, no. 4, pp. 1259–1267, 2013. [30] N. J. Gusani, F. K. Balaa, J. L. Steel et al., “Treatment of unresectable cholangiocarcinoma with gemcitabine-based transcatheter arterial chemoembolization (TACE): a single- institution experience,” Journal of Gastrointestinal Surgery, vol. 12, no. 1, pp. 129–137, 2008. [31] T. M. Shaker, C. Chung, M. K. Varma et al., “Is there a role for ytrrium-90 in the treatment of unresectable and metastatic intrahepatic cholangiocarcinoma?” 6e American Journal of Surgery, vol. 215, no. 3, pp. 467–470, 2018. [32] A. Cucchetti, A. Cappelli, C. Mosconi et al., “Improving patient selection for selective internal radiation therapy of

Journal

Journal of OncologyHindawi Publishing Corporation

Published: Oct 10, 2020

There are no references for this article.