If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author. Scientific Bureau, Dutch Institute of Clinical Auditing, 2333 AA, Leiden, the Netherlands.
Affiliations
Dutch Institute for Clinical Auditing, Scientific Bureau, Leiden, the NetherlandsDepartment of Surgery, University Medical Centre Groningen, Groningen, the Netherlands
Routine treatment with preoperative systemic chemotherapy (CTx) in patients with colorectal liver metastases (CRLM) remains controversial due to lack of consistent evidence demonstrating associated survival benefits. This study aimed to determine the effect of preoperative CTx on overall survival (OS) compared to surgery alone and to assess hospital and oncological network variation in 5-year OS.
Methods
This was a population-based study of all patients who underwent liver resection for CRLM between 2014 and 2017 in the Netherlands. After 1:1 propensity score matching (PSM), OS was compared between patients treated with and without preoperative CTx. Hospital and oncological network variation in 5-year OS corrected for case-mix factors was calculated using an observed/expected ratio.
Results
Of 2820 patients included, 852 (30.2%) and 1968 (69.8%) patients were treated with preoperative CTx and surgery alone, respectively. After PSM, 537 patients remained in each group, median number of CRLM; 3 [IQR 2–4], median size of CRLM; 28 mm [IQR 18–44], synchronous CLRM (71.1%). Median follow-up was 80.8 months. Five-year OS rates after PSM for patients treated with and without preoperative chemotherapy were 40.2% versus 38.3% (log-rank P = 0.734). After stratification for low, medium, and high tumour burden based on the tumour burden score (TBS) OS was similar for preoperative chemotherapy vs. surgery alone (log-rank P = 0.486, P = 0.914, and P = 0.744, respectively). After correction for non-modifiable patient and tumour characteristics, no relevant hospital or oncological network variation in five-year OS was observed.
Conclusion
In patients eligible for surgical resection, preoperative chemotherapy does not provide an overall survival benefit compared to surgery alone.
Surgical resection is considered the only potential cure for colorectal liver metastases (CRLM) and provides 5-year overall survival (OS) rates of approximately 50% in selected patients [
Impact of patient, primary tumor and metastatic pattern including tumor location on survival in patients undergoing ablation or resection for colorectal liver metastases: a population-based national cohort study.
Rates and patterns of recurrence following curative intent surgery for colorectal liver metastasis: an international multi-institutional analysis of 1669 patients.
]. Recent improvements in operative techniques combined with more effective chemotherapy (CTx) increased the proportion of patients eligible for surgical resection, thereby improving survival rates [
Improved long-term outcome of surgery for advanced colorectal liver metastases: reasons and implications for management on the basis of a severity score.
Treatment strategies in colorectal cancer patients with initially unresectable liver-only metastases, a study protocol of the randomised phase 3 CAIRO5 study of the Dutch Colorectal Cancer Group (DCCG).
]. The EORTC 40983 trial reported an improved disease-free survival (DFS) but without an OS benefit in patients with resectable CRLM treated with perioperative CTx [
Perioperative FOLFOX4 chemotherapy and surgery versus surgery alone for resectable liver metastases from colorectal cancer (EORTC 40983): long-term results of a randomised, controlled, phase 3 trial.
]. More recently, these results were confirmed for adjuvant chemotherapy by the findings of the JCOG0603 trial reporting an improved DFS but no OS benefit in patients treated with adjuvant CTx compared to upfront surgery [
Hepatectomy followed by mFOLFOX6 versus hepatectomy alone for liver-only metastatic colorectal cancer (JCOG0603): a phase II or III randomized controlled trial.
]. These studies were interpreted differently by various countries and even within countries.
Generalising these results to all patients with resectable metastases is difficult as both studies mainly included patients with a single liver metastasis, typically representing patients with more favourable prognoses [
Perioperative FOLFOX4 chemotherapy and surgery versus surgery alone for resectable liver metastases from colorectal cancer (EORTC 40983): long-term results of a randomised, controlled, phase 3 trial.
Hepatectomy followed by mFOLFOX6 versus hepatectomy alone for liver-only metastatic colorectal cancer (JCOG0603): a phase II or III randomized controlled trial.
]. With reference to the results of preceding RCTs, advocates of preoperative CTx will emphasize the improved DFS, whereas opponents of preoperative CTx will emphasize the lack of OS benefit. As a result, there is considerable practice variation in the administration of preoperative chemotherapy [
The aim of this nationwide population-based study was to assess if preoperative CTx was associated with improved 5-year overall survival compared to surgery alone stratified to tumour burden score. Additionally, the variation between hospitals and oncology networks in 5-year OS was assessed.
2. Methods
This nationwide, population-based study was conducted with data from the Dutch Hepato Biliary Audit (DHBA) registry, Vektis, and the Dutch Municipal Personal Records Database (BPR). Since 2013, the DHBA is a compulsory clinical audit that registers all patients undergoing surgery with the intent of liver resection or thermal ablation [
]. In the Netherlands, hospitals are required to perform a minimum of 20 liver resections per year. Seven oncological networks exist, these include at least one tertiary referral centre and several regional hospitals. Oncological networks have been established to optimize collaboration and decrease variation [
]. The DHBA scientific committee approved this study. No ethical approval or informed consent was needed under the Dutch law, as data was handled anonymously.
The DHBA does not contain long-term follow-up data. Therefore, the DHBA was combined with the Vektis and BPR databases to assess overall survival. Vektis is the Dutch national claim database for health insurance companies and receives the date of death of all the deceased with healthcare insurance from the BPR, covering over 99% of Dutch inhabitants. However, in three hospitals (10.2%) linkage with the Vektis database was not possible; therefore, for assessment of overall survival in patients treated in these hospitals, DHBA data was directly linked with the BPR database. Data of colorectal resections are registered in a separate audit (Dutch Colorectal Audit). Merging of data is prohibited due to general data protection regulation (GDPR). This means that data on the site of the colon tumour (left vs. right), and pathological T stage are missing. N stage was requested in the DHBA if available.
Datasets were combined based on unique personal citizen service numbers. To guarantee privacy, linkage of datasets was performed by a third-trusted party Medical Research Data Management (MRDM). After linkage of the datasets, follow-up data of 99.3% of the patients in the DHBA was covered. Patients not registered as deceased at the time of merging the datasets (2022, July) were assumed to be alive.
Overall survival was calculated from the date of surgery to the date of death of any cause. Date of start of preoperative CTx is not registered in the DHBA.
2.1 Patients
All patients of ≥18 years old, who underwent liver resection for CRLM between 1st of January 2014 and the 31st of December 2017 were included in this study. Patients with a history of liver resection or treated with ablation alone were excluded. Furthermore, patients with missing essential data on preoperative chemotherapy or the type of tumour treatment were excluded. For analyses of overall survival, patients with in-hospital or 30-day mortality were excluded.
At the time included patients were operated on, the Dutch guidelines for (metastatic) colorectal cancer disclosed no clear preference for perioperative chemotherapy. For patients with primarily resectable metastases chemotherapy was not considered standard treatment, yet definitions of resectability were lacking. This contributed to practice variation between centres and networks [
Population-based study on practice variation regarding preoperative systemic chemotherapy in patients with colorectal liver metastases and impact on short-term outcomes.
]. Adjuvant chemotherapy was and is not supported and therefore rare as opposed to many countries where it is common practice.
2.2 Treatment groups
For analyses, patients were divided into two treatment groups: patients who received preoperative CTx and liver resection and those who underwent surgery without receiving preoperative CTx. Preoperative chemotherapy was defined as any form (e.g., neoadjuvant or induction therapy) of CTx administrated as a treatment for CRLM prior to surgery. It is difficult to distinguish between true neo-adjuvant or induction therapy, and possibly these definitions overlap. This results from nuances in the presentation of the individual patient, which can blur the boundaries of what is meant by upfront resectable [
In a subsequent part of the patients registered in the DHBA, indication for preoperative chemotherapy was reported (as neoadjuvant or induction), planned additional analyses for this share of patients on OS were performed to account for possible bias by indication for preoperative chemotherapy.
2.3 Variables for analyses
Patient characteristics used for analysis included sex, age, body mass index (BMI), American Society of Anaesthesiologists (ASA) grade, and Charlson Comorbidity Index (CCI). Tumour characteristics used for analysis included number of CRLM, diameter of the largest CRLM before tumour-specific treatment, synchronous or metachronous metastases, CEA, pathological N-stage of primary tumour, location of primary tumour (rectal/colon), bilobar disease, and extrahepatic disease. In the registry location within the colon was not specified (i.e., left-sided or right-sided). Treatment characteristics used for analysis included surgical procedure (open/laparoscopic), major liver resection (resection of ≥3 adjacent Couinaud segments), and type of hospital where treatment took place (tertiary referral centre or regional hospital).
2.4 Statistical analyses
Categorical variables were presented as frequencies and proportions. In the unmatched and matched cohort, the significance of variables was tested using the Chi-square test or Fisher's exact test for categoric variables and the student's t-test for continuous data.
To compare outcomes of patients treated with or without preoperative chemotherapy and reduce baseline differences in these groups propensity score matching (PSM) was performed. Propensity scores were obtained from a logistic regression model. Variables used to calculate propensity scores included age, ASA score, extrahepatic disease, location of primary tumour, timing of CRLM (synchronous vs. metachronous), number of CRLM, diameter of largest CRLM before start of tumour specific treatment, and major liver resection. N-stage of primary tumour and CEA-level were not used as variables in PSM analysis since over 25% data was missing. This is because no information was available when colorectal resections were performed after the liver resection. PSM was performed with a 1:1 ratio using the nearest neighbour method with a calliper of 0.07. All patients without a matching counterpart were excluded from the analyses. Balance after PSM was assessed using standardised mean difference (SMD). SMD <0.10 indicates an optimal balance of confounding factors.
Kaplan-Meier survival analyses with the log-rank test were used to compare overall survival between patients treated with preoperative CTx and surgery alone. A planned subgroup analysis was performed in the original unmatched and matched cohort for three zones of the Tumour Burden Score (TBS). The TBS is calculated as (TBS)2 = (diameter of largest CRLM in cm)2 + (number of CRLM)2. The TBS is stratified into a low (TBS of 0–2), medium (TBS of 3–8) and high (TBS ≥9) tumour burden [
Between hospital and oncology network variation in five-year OS corrected for case-mix factors was calculated using an observed/expected ratio. In this O/E ratio, the authors accounted for the possibility of censoring since this is one of the most distinguished features of survival data. The observed (O) was calculated by the number of events for each patient at the specified follow-up period. The expected (E) was the number of events expected in a centre or network, based on the number of patients, their follow-up, and their patient and tumour characteristics, when the centre performed according to the benchmark. The benchmark was determined by fitting a single Cox model to the complete data of all centres, representing the average performance of all centres. More details are described previously [
]. An O/E ratio above 1 indicated that a hospital's or network's five-year OS was worse than expected. The 95% confidence intervals were calculated to indicate whether a hospital's or network's performance was statistically different from the other hospitals.
All analyses were performed using R version 4.2.0 (R Core Team (2021). (R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria).
3. Results
A total of 2820 patients were included in this study (Fig. 1). Of these patients, 852 (30.2%) received preoperative CTx, and 1968 (69.8%) were treated with surgery alone.
Patients treated with preoperative CTx were younger, had more CRLM, higher TBS, greater diameter of CRLM, and more often had synchronous metastases, bilobar metastases, and extrahepatic disease compared to patients who underwent surgery alone (Table 1). In the surgery-alone group none of the patients received treatment with adjuvant CTx.
Table 1Baseline characteristics of patients with CRLM treated with surgery alone, or preoperative systemic chemotherapy and surgery between 2014 and 2017. Before and after propensity score matching.
Bold p-values indicates a statistical significant values (p < 0.05). Abbreviations: Standardised mean differences (SMD), BMI indicates body mass index; ASA score indicates American Association of Anaesthesiologist; type of hospital tertiary centre indicates hospitals with the highest expertise on oncological surgery.
After propensity score matching, 537 patients remained in each group. SMD was below 0.10 for most patients and tumour characteristics, except histopathology of the liver (P < 0.001, SMD = 0.288), N-stage of the primary tumour (P = 0.080, SMD = 0.173) and CEA-level (P = 0.019, SMD = 0.173) (Table 1).
3.2 Treatment effects of preoperative chemotherapy on overall survival
The median follow-up time in the preoperative CTx group was 84.3 months (95% CI 73.3–95.4) versus 83.6 months (95% CI 73.9–96.7) in the surgery alone group. Five-year OS rates before matching for preoperative CTx versus surgery alone were 49.1% versus 36.9% (P < 0.001). Within the original unmatched cohort, median survival in the preoperative CTx cohort was 39.9 months (95% CI 35.9–45.1 versus 58.1 months (95% CI 54.0–63.9) in the surgery alone cohort.
After PSM, median follow-up time in the preoperative CTx cohort was 83.3 months (95% CI 72.9–92.7) versus 82.6 months (95% CI 73.2–93.9) in the surgery alone group. Five-year OS for preoperative CTx versus surgery alone was 43.3% versus 42.6% (P = 0.734), respectively (Fig. 2). In the matched cohort, the median survival time in the preoperative CTx cohort was 43.3 months (95% CI 37.7–47.9) versus 42.6 months (95% CI 37.3–48.2) in the surgery alone cohort. Additional OS analyses to account for possible bias by indication of preoperative chemotherapy showed no differences in OS between treatment groups (Supplementary Figs. 1 and 2). The purpose of chemotherapy (i.e., neoadjuvant or induction) was clearly stated in the patients included in these analyses.
Fig. 2Kaplan – Meier survival curves presenting OS of patients with CRLM receiving preoperative systemic chemotherapy and surgery or surgery alone between 2014 and 2017. A. OS for surgery alone (blue) and surgery and preoperative CTx (grey) before matching B. OS after PSM for surgery alone (blue) and surgery and preoperative CTx (grey). Time in months.
3.3 Treatment effects of preoperative chemotherapy on overall survival in sub-groups based on tumour burden score (TBS)
After stratification for TBS score, preoperative chemotherapy did not improve overall survival. In the unmatched cohort five-year OS in the TBS low group was 52.9% in the preoperative CTx group and 59.0% in the surgery alone group, (P = 0.0160). For TBS medium, and TBS high burden five-year OS rates were, respectively, 37.2% vs. 42.6% (P = 0.039), 31.9% vs. 31.9% (P = 0.914) (Fig. 3).
Fig. 3Kaplan – Meier survival curves presenting OS of patients with CRLM receiving preoperative CTx and surgery alone between 2014 and 2017 stratified for Tumour Burden Score (TBS) in the unmatched cohort. A: the unmatched cohort, B: PS matched cohort. 1. OS for preoperative CTx (grey) and surgery alone (blue) after stratification for TBS 0–2. 2. OS for preoperative CTx (grey) and surgery alone (blue) after stratification for TBS 3–8. 3. OS for preoperative CTx (grey) and surgery alone (blue) after stratification for TBS ≥9. Time in months.
Five-year OS for TBS low, medium and high group after PSM were 54.8%, 37.3% and 35.6% in the preoperative CTx group, 49.4%, 36.9% and 31.6% in the surgery alone group (P = 0.486, P = 0.914, and P = 0.744), respectively. Median survival time after PSM for TBS low, medium and high were 75.2 months (95% CI 49.8 - NA), 39.3 months (95% CI 33.0–45.1) and 40.8 months (95% CI 35.0–62.9) in the preoperative CTx group and 59.7 months (95% CI 47.1 - NA), 39.7 months (95% CI 34.5–46.8), and 36.1 months (95% CI 21.1 – NA) in the surgery alone group, respectively.
3.4 Synchronous versus metachronous CRLM
After PSM, no difference in overall survival was observed in patients treated with and without preoperative CTx in subgroups of synchronous CRLM (p = 0.55) and metachronous (p = 0.13) CRLM, respectively (Supplementary Figs. 4 and 5).
3.5 Hospital and oncology network variation in OS
Unadjusted hospital variance in OS ranged between 0.67 and 1.20 (Supplementary Fig. 3). Unadjusted oncological network variance ranged between 0.95 and 1.04 (Supplementary Fig. 6A). After correction for case-mix factors, adjusted hospital variance ranged between 0.68 and 1.26 (Fig. 4), and adjusted network variance ranged between 0.97 and 1.06 (Supplementary Fig. 6B). No hospital or oncological network fell out of the 95% confidence interval of variance for five-year OS.
Fig. 4Benchmarked five-year OS between hospital in the Netherlands for patients with CRLM between 2014–201. Hospital variation in five-year OS between 2014 and 2017 corrected for case-mix factors∗
∗Case-mix correction for the following variables: Age, sex, BMI, ASA score, Charlson Comorbidity score, Number of CRLM, Diameter of largest CRLM, major liver resection, Timing of disease, Extrahepatic disease, location of primary tumour.
In this analysis of real-world data preoperative chemotherapy did not improve survival as compared to surgery alone. In sub-analyses in which patients were divided according to tumour burden scores or recorded intention of treatment (induction or neoadjuvant) preoperative chemotherapy did not prolong overall survival in patients either. Variation between hospitals and networks in use of preoperative chemotherapy did not lead to variation in overall survival after correction for case-mix.
Randomised Controlled Trials (RCTs) represent the gold standard in the assessment of the efficacy of a therapy, but a major criticism is that results cannot always be extrapolated to the real-world setting due to strict selection criteria [
]. Results of this real-world data study were concordant with important RCTs (EORTC 40983 trial and JCOG0603 trial), describing no OS benefit in patients treated with perioperative chemotherapy or surgery alone [
Perioperative FOLFOX4 chemotherapy and surgery versus surgery alone for resectable liver metastases from colorectal cancer (EORTC 40983): long-term results of a randomised, controlled, phase 3 trial.
Hepatectomy followed by mFOLFOX6 versus hepatectomy alone for liver-only metastatic colorectal cancer (JCOG0603): a phase II or III randomized controlled trial.
]. Unfortunately, results on DFS could not be confirmed since the DHBA does not contain any information on disease-free survival of all registered patients. Moreover, Ecker et al. described a poor correlation between disease-free survival and overall survival in patients with CRLM since development of recurrent CRLM after liver resection did not reflect non-curability [
Recurrence-free survival versus overall survival as a primary endpoint for studies of resected colorectal liver metastasis: a retrospective study and meta-analysis.
In theory, administration of neoadjuvant chemotherapy in patients with resectable CRLM could reduce risk of recurrence by treatment of micro-metastatic disease, also it could downsize the CRLM and improve obtaining negative resection margins [
Rates and patterns of recurrence following curative intent surgery for colorectal liver metastasis: an international multi-institutional analysis of 1669 patients.
Recurrence-free survival versus overall survival as a primary endpoint for studies of resected colorectal liver metastasis: a retrospective study and meta-analysis.
]. Previous studies found no differences in intrahepatic recurrent disease in patients treated with or without perioperative chemotherapy, but they observed an overall survival benefit in high-risk patients [
]. Prognosis of patients with recurrent disease is favourable if local treatment of recurrent metastases is possible. This could explain why earlier mentioned improved DFS in patients treated with preoperative chemotherapy did not translate into improved OS [
Several small observational studies suggested a potential survival benefit of administering preoperative chemotherapy to patients with high-risk of recurrence [
Neoadjuvant chemotherapy versus upfront surgery for resectable liver metastases from colorectal cancer: a multicenter, propensity score-matched cohort study.
Clinical outcomes of neoadjuvant chemotherapy in colorectal cancer patients with synchronous resectable liver metastasis: a propensity score matching analysis.
]. The current study accounted for ‘low-, medium- and high-risk patients' by stratification based on TBS and did not observe a beneficial effect of preoperative CTx within these groups [
]. The discrepancy in described survival benefits of preoperative CTx in high-risk patients could be attributed to the absence of uniform criteria to define high risk; different risk scores were used in literature based on various clinicopathological and biological factors. Clinical risk scores have been validated to predict overall survival in patients, but have not been validated to guide decisions in the management of CRLM, such as treatment with preoperative chemotherapy [
]. Moreover, decision on which patients need to be treated with preoperative CTx probably depends more on the clinical experience and personal preference of multidisciplinary teams rather than outcomes of scoring systems only [
Neoadjuvant chemotherapy versus upfront surgery for resectable liver metastases from colorectal cancer: a multicenter, propensity score-matched cohort study.
Clinical outcomes of neoadjuvant chemotherapy in colorectal cancer patients with synchronous resectable liver metastasis: a propensity score matching analysis.
Determined by previous studies, various results and interpretations on the use of preoperative chemotherapy in resectable CRLM could influence the different personal preferences of multidisciplinary teams. Consequently, practice variation occurs. Elfrink et al. reported significant practice variation (2%–55%) regarding the use of preoperative CTx between Dutch hospitals [
Population-based study on practice variation regarding preoperative systemic chemotherapy in patients with colorectal liver metastases and impact on short-term outcomes.
]. However, this did not result in varying survival between hospitals or oncological networks before and after adjustment for case-mix factors. These results, can be interpreted as reassuring as patients seem to have a rather similar chance of survival wherever they are treated in the Netherlands. Caution is needed, however, as residual confounding cannot be ruled out. Only the patients who finally underwent resection were included.
The authors advocate that routine use of chemotherapy in patients with upfront, clearly resectable, CRLM should stop [
Perioperative FOLFOX4 chemotherapy and surgery versus surgery alone for resectable liver metastases from colorectal cancer (EORTC 40983): long-term results of a randomised, controlled, phase 3 trial.
Hepatectomy followed by mFOLFOX6 versus hepatectomy alone for liver-only metastatic colorectal cancer (JCOG0603): a phase II or III randomized controlled trial.
]. They would like to emphasize that this does not mean that preoperative chemotherapy should never been given. When downstaging of the primary tumour or metastases is needed, there is a clear indication to start with chemotherapy [
]. When oncologically unfavourable characteristics such as extrahepatic disease or multiple metastases are present, chemotherapy is likely needed. The main aim of chemotherapy here is to rule out patients with rapidly progressive and unfavourable metastatic disease and, therefore, would not benefit from a potential morbid surgical resection. The international HPB community should put effort into determining these high-risk patients.
This study was unique by the fact that it used real-world (audit) data of all patients who underwent liver resection, reflecting daily practice in the Netherlands. These results complement the evidence from RCTs with information about use or preoperative chemotherapy on a broader population. Limitations of the present study include the use of clinical audit data. Important details, including the decision to administer chemotherapy of the multidisciplinary team (MDT), exact reason for chemotherapy, start date, type, and number of cycles of chemotherapy, were lacking. PSM was used to balance groups but could not rule out residual confounding. Patients who were not eligible for liver resection after chemotherapy were missing. The DHBA did not register any data on recurrence, meaning disease-free survival could not be analysed. As expected, few patients were available for analysis of the effect of preoperative chemotherapy on survival in the high tumour burden score group. Most of these patients did get systemic treatment first, which aligns with the common perception that patients with the highest risk of recurrence need perioperative chemotherapy. Unfortunately, several attempts to study this high-risk group in a randomised settings failed [
Neo-adjuvant chemotherapy followed by surgery versus surgery alone in high-risk patients with resectable colorectal liver metastases: the CHARISMA randomized multicenter clinical trial.
]. The current study could not provide enough evidence to answer whether patients with high TBS scores and resectable CRLM should get preoperative chemotherapy. These data have been included in a national meeting of liver surgeons in the Netherlands. The authors are keen to observe a decline in the use of preoperative chemotherapy in these patients in the Netherlands over the following years.
5. Conclusion
In conclusion, this study did not observe an overall survival benefit for patients who received preoperative chemotherapy compared to surgery alone in real-world data. Therefore, routine use of preoperative CTx should not be recommended in patients with low and medium TBS who are eligible for surgical resection.
CRediT authorship contribution statement
Michelle.R. de Graaff: Conceptualization, Data curation, Formal analysis, Writing – original draft. Joost M. Klaase: Conceptualization, Data curation, Writing – review & editing, Formal analysis, Supervision. Ronald M. van Dam: Conceptualization, Data curation, Writing – review & editing. Koert F.D. Kuhlmann: Conceptualization, Data curation, Writing – review & editing. Geert Kazemier: Conceptualization, Data curation, Writing – review & editing. Rutger-Jan Swijnenburg: Conceptualization, Data curation, Writing – review & editing. Arthur K.E. Elfrink: Conceptualization, Data curation, Formal analysis, Writing – review & editing. Cees Verhoef: Conceptualization, Data curation, Writing – review & editing. J.Sven Mieog: Conceptualization, Data curation, Writing – review & editing. Peter B. van den Boezem: Conceptualization, Data curation, Writing – review & editing. Paul Gobardhan: Conceptualization, Data curation, Writing – review & editing. Arjen M. Rijken: Conceptualization, Data curation, Writing – review & editing. Daan J. Lips: Conceptualization, Data curation, Writing – review & editing. Wouter G.K. leclercq: Conceptualization, Data curation, Writing – review & editing. Hendrik A. Marsman: Conceptualization, Data curation, Writing – review & editing. Peter van Duijvendijk: Conceptualization, Data curation, Writing – review & editing. Joost A.B. van der Hoeven: Conceptualization, Data curation, Writing – review & editing. Maarten Vermaas: Conceptualization, Data curation, Writing – review & editing. Marcel den Dulk: Conceptualization, Data curation, Writing – review & editing, Formal analysis, Supervision. Dirk J. Grünhagen: Conceptualization, Data curation, Writing – review & editing, Formal analysis, Supervision. Niels F.M. Kok: Conceptualization, Data curation, Writing – review & editing, Formal analysis, Supervision. Carlijn I. Buis: Conceptualization, Data curation, Formal analysis, Writing – review & editing. Jeroen Hagendoorn: Conceptualization, Data curation, Writing – review & editing. Wouter J.M. Derksen: Conceptualization, Data curation, Writing – review & editing. Hans Torrenga: Conceptualization, Data curation. Eric Manusama: Conceptualization, Data curation, Writing – review & editing. N. Tjarda van Heek: Conceptualization, Data curation, Writing – review & editing. Steven J. Oosterling: Conceptualization, Data curation, Writing – review & editing. Koop Bosscha: Conceptualization, Data curation, Writing – review & editing. Andries E. Braat: Conceptualization, Data curation, Writing – review & editing. Frederik J.H. Hoogwater: Conceptualization, Data curation, Writing – review & editing. Esther C.J. Consten: Conceptualization, Data curation, Writing – review & editing. Christiaan van der Leij: Conceptualization, Data curation, Writing – review & editing. Mark C. Burgmans: Conceptualization, Data curation, Writing – review & editing. Mike S.L. Liem: Conceptualization, Data curation, Writing – review & editing. Eric J.Th Belt: Conceptualization, Data curation, Writing – review & editing. Gijs A. Patijn: Conceptualization, Data curation, Writing – review & editing.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
The authors would like to thank all surgeons, interventional radiologists, and administrative nurses for data registration in the DHBA database, as well as the Dutch Hepato Biliary Audit Group for scientific input.
Appendix A. Supplementary data
The following are the Supplementary data to this article:
Kaplan-Meier curves to asses overall survival in patients registered as neo-adjuvant CTx.
A: Full cohort. OS for neo-adjuvant CTx (grey) and surgery alone (blue), time in days B: After PS matching. OS for neo-adjuvant CTx (grey) and surgery alone (blue), time in months
Kaplan-Meier curves to asses overall survival in patients registered as induction CTx.
A: Full cohort. OS induction CTx (grey) and surgery alone (blue), time in days B: After PS matching. OS for induction CTx (grey) and surgery alone (blue), time in months
Kaplan-Meier curves to asses overall survival in patients with synchronous CRLM. A: Full cohort. OS for neo-adjuvant CTx (grey) and surgery alone (blue), time in days B: After PS matching. OS for neo-adjuvant CTx (grey) and surgery alone (blue), time in months
Kaplan-Meier curves to asses overall survival in patients with metachronous CRLM.
A: Full cohort. OS for neo-adjuvant CTx (grey) and surgery alone (blue), time in days B: After PS matching. OS for neo-adjuvant CTx (grey) and surgery alone (blue), time in months
Five-year OS between oncological networks in the Netherlands for patients with CRLM between 2014 and 2017. A: uncorrected oncological network variation in five-year OS B: Oncological network variation in five-year OS between 2014 and 2017 corrected for case-mix factors∗
∗Case-mix correction for the following variables: Age, sex, BMI, ASA score, Charlson Comorbidity score, Number of CRLM, Diameter of largest CRLM, major liver resection, Timing of disease, Extrahepatic disease, location of primary tumour.
References
Scherman P.
Syk I.
Holmberg E.
Naredi P.
Rizell M.
Impact of patient, primary tumor and metastatic pattern including tumor location on survival in patients undergoing ablation or resection for colorectal liver metastases: a population-based national cohort study.
Rates and patterns of recurrence following curative intent surgery for colorectal liver metastasis: an international multi-institutional analysis of 1669 patients.
Improved long-term outcome of surgery for advanced colorectal liver metastases: reasons and implications for management on the basis of a severity score.
Treatment strategies in colorectal cancer patients with initially unresectable liver-only metastases, a study protocol of the randomised phase 3 CAIRO5 study of the Dutch Colorectal Cancer Group (DCCG).
Perioperative FOLFOX4 chemotherapy and surgery versus surgery alone for resectable liver metastases from colorectal cancer (EORTC 40983): long-term results of a randomised, controlled, phase 3 trial.
Hepatectomy followed by mFOLFOX6 versus hepatectomy alone for liver-only metastatic colorectal cancer (JCOG0603): a phase II or III randomized controlled trial.
Population-based study on practice variation regarding preoperative systemic chemotherapy in patients with colorectal liver metastases and impact on short-term outcomes.
Recurrence-free survival versus overall survival as a primary endpoint for studies of resected colorectal liver metastasis: a retrospective study and meta-analysis.
Neoadjuvant chemotherapy versus upfront surgery for resectable liver metastases from colorectal cancer: a multicenter, propensity score-matched cohort study.
Clinical outcomes of neoadjuvant chemotherapy in colorectal cancer patients with synchronous resectable liver metastasis: a propensity score matching analysis.
Neo-adjuvant chemotherapy followed by surgery versus surgery alone in high-risk patients with resectable colorectal liver metastases: the CHARISMA randomized multicenter clinical trial.
00491-2&id=gr1.jpg){kind=link}
00491-2&id=gr2.jpg){kind=link}
00491-2&id=gr3.jpg){kind=link}
00491-2&id=gr4.jpg){kind=link}