Primary tumor resection with or without metastasectomy for left- and right-sided stage IV colorectal cancer: an instrumental variable analysis
BMC Gastroenterology volume 22, Article number: 114 (2022)
Whether primary tumor location (PTL) is predictive of survival benefits following primary tumor resection plus metastasectomy (PMTR) and primary tumor resection (PTR) alone in stage IV colorectal cancer patients is not known. We sought to address this issue by employing instrumental variable analysis to evaluate the efficacy of PMTR and PTR with stratification for primary tumor location in stage IV colorectal cancer patients.
Patients and methods
Stage IV colorectal cancer patients diagnosed between January 1, 2005 and December 31, 2015 were identified from the Surveillance, Epidemiology, and End Results Program of the National Cancer Institute. To account for both measured and unmeasured confounders, the efficacy of PMTR and PTR in the left- and right-sided subgroups was evaluated using instrumental variable analysis, with the health service area as the instrument variable. Overall survival (OS) was the primary outcome of interest.
A total of 50,333 eligible patients were analyzed (left-sided, n = 29,402 and right-sided, n = 20,931). OS was significantly better with PMTR than with other treatments (PTR, metastasectomy only, or no surgery) in patients with left-sided tumors (hazard ratio [HR] = 0.37 [95% CI 0.24–0.58], P < 0.001), but not in patients with right-sided tumors (HR = 0.98 [95% CI 0.65–1.47], P = 0.910; interaction test P < 0.001). OS was comparable in patients treated with PTR and those treated with no surgery in both the left-sided (HR = 1.11 [95% CI 0.68–1.81], P = 0.690) and right-sided (HR = 0.85 [95% CI 0.50–1.43], P = 0.530; interaction test P = 0.466) subgroups.
PMTR appears to only benefit patients with left-sided stage IV colorectal cancer but not those with right-sided tumors. PTR does not improve OS, regardless of primary tumor location. When selecting patients for PMTR, primary tumor location should be considered. Overuse of PTR should be avoided.
Increased use of primary tumor resection plus metastasectomy benefited patients with left-sided tumors but not those with right-sided tumors.
Increased use of primary tumor resection alone did not improve survival in either left or right-sided tumor group.
Overuse of palliative primary tumor resection should be avoided.
According to GLOBOCAN, colorectal cancer (CRC) accounted for 10% of all cancers and for 9% of all cancer deaths in 185 countries in 2018 . Approximately 20% of CRC patients have distant metastases at the time of diagnosis . In those with resectable primary tumor and synchronous metastases, staged or simultaneous primary tumor resection plus metastasectomy (PMTR) leading to no evidence of disease (NED) can improve long-term survival . Previous study found metastasectomy would improve OS in left-sided colon cancer but not in right-sided colon cancer . However, there is still a lack of consensus on the resectability of metastatic lesions, and previous population-based studies have reported a potential underutilization of PMTR among stage IV CRC patients [5, 6]. According to the National Comprehensive Cancer Network (NCCN) Clinical Practice Guidelines, patients with symptomatic primary tumor (i.e., with obstruction, bleeding, and so on) should receive primary tumor resection (PTR), but those with asymptomatic metastatic CRC (stage IV CRC) should receive systemic chemotherapy as the initial treatment as it is still uncertain whether PTR improves outcomes in this group [7,8,9,10,11,12,13]. Recently, a randomized trial has demonstrated PTR is not associated with survival benefit in asymptomatic patients . Although there is a decline in the number of PTRs being performed, it continues to be overused in clinical practice .
Depending on the differences in embryonic origin, the colon had divided into right-sided and left-sided. Right-sided CRC occurs in cecum, ascending colon, hepatic flexure and transverse colon, while left-sided CRC occurs in splenic flexure, descending colon, sigmoid colon and rectal. Accumulating evidence has revealed that tumor arising from different sides of the colon exhibit different features, such as epidemiological incidence, physiological characteristics, molecular alterations, and even survival outcome [16,17,18]. PTL was found to be a independent prognostic factor in stage IV CRC patients treated with systemic therapy and patients with left-sided tumors had a significantly better prognosis than those with right-sided tumors . Price et al. study have shown that overall survival (OS) of stage IV CRC receiving hepatic surgery with or without chemotherapy had no significant difference by site, but OS of patients receiving hepatic chemotherapy only had markedly improved in patients with left-sided than those with right-sided . So far, however, most of PTR or PMTR of stage IV CRC studies were based on all CRC populations, there have been no studies examining how PTL affects the outcomes of PMTR or PTR. The survival benefit of PMTR and PTR for site-specific stage IV CRC remains unclear.
Instrumental variable (IV) analysis, first introduced by Brookhart from econometrics to observational study , is a method useful in controlling both the measured and potential unmeasured confounders and to estimate causal-effect of treatment by relying on natural variation in treatment choice. In this study, we employed IV analysis to determine whether PTL affects survival outcomes of PMTR or PTR in stage IV CRC patients.
Patient eligibility and exclusion criteria
This population-based study used data extracted from the November 2018 ASCI text–data version of the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute, which covers ~ 28% of all cancer cases in the US . We identified patients with CRC using the International Classification of Diseases for Oncology (ICD-O-3) site codes C180, C182 to C189, C199, and C209. We only considered patients with adenocarcinoma identified by the ICD-O-3 histology codes 8140, 8144, 8210, 8211, 8220, 8221, 8261, 8262, 8263, 8480, and 8481. Stage IV disease was identified based on the North American Association of Central Cancer Registries (NAACCR) items 790, 810, and 2980. We only included patients diagnosed with CRC in and after year 2005—i.e., after the two landmark targeted therapeutic agents cetuximab and bevacizumab were approved for use in metastatic CRC by the US Food and Drug Administration . We excluded patients who (1) were < 18 years old; (2) did not have histologically confirmed diagnosis; (3) had previous history of other primary malignancy; (4) had been diagnosed only after death (e.g., at autopsy); (5) had incomplete follow-up information; or (6) had incomplete information on surgery to the primary or metastatic sites.
Study endpoint and the examined demographic and clinical variables
OS—defined as the time from diagnosis to the date of death due to any cause—was the primary outcome of interest. The examined covariates included race, age, sex, marital status, year of diagnosis, health service area (HSA), tumor grade, tumor location, tumor size, American Joint Committee on Cancer (AJCC) T and N category, surgery to the primary site, and surgery to the metastatic site. Following the criteria used in prior studies, patients with cancer of the cecum, ascending colon, and transverse colon were defined as having right-sided tumors, and patients with cancer of the splenic flexure, descending colon, sigmoid colon, rectosigmoid junction, and rectum were considered as having left-sided tumors .
This study is a population-based cohort study extracted from the Surveillance, Epidemiology, and End Results Program of the National Cancer Institute with stage IV colorectal cancer patients diagnosed between January 1, 2005 and December 31, 2015. To account for both measured and unmeasured confounders, the efficacy of PMTR and PTR in the left- and right-sided subgroups was evaluated using instrumental variable analysis, with the health service area as the instrument variable. Overall survival (OS) was the primary outcome of interest. Individual consent for this retrospective analysis was waived.
We applied instrumental variable analytical methods, which can account for both measured and unmeasured confounders. Most of observational studies cannot achieve the inference of causality because the limitations and bias of design of studies (especially the unmeasured confounders). The unmeasured confounders are associated to exposure and outcome, which are important to drawing causality in observational studies [25,26,27,28].
IV requires one or more instruments, which are associated with exposure but are not associated with other confounding factors, and has no directly relationship with outcome variable. Two-stage predictor substitution (2SPS) and two-stage residual inclusion (2SRI) are two commonly used IV approaches, which are used in nonlinear models. We chose 2RSI for the instrumental variable analysis rather than 2SPS because the former was more consistent and less biased for solving problem of nonlinear models endogeneity [29, 30].
In contrast to a randomized controlled trial that identifies the average treatment effect [31, 32], instrumental variable analysis estimates the treatment effect for the marginal patients, i.e., the patients in whom the likelihood of undergoing the treatment depends on the instrumental variable . The HSA—defined as one or more counties that are relatively self-contained with respect to the provision of routine hospital care —was adopted as the instrumental variable because the difference in the use of PTR and PMTR across different HSAs is more likely to be associated with nonmedical factors (e.g., local treatment practices) than with differences in cancer characteristics. Each HSA with fewer than 50 cases was combined with the nearest (in terms of geographic distance) HSA. We used the Durbin–Wu–Hausman test of endogeneity to decide on the necessity of performing instrumental variable analysis; a significant result indicated that standard multivariable regression was likely to provide biased results .
Instrumental variable analysis was performed separately in the left-sided and right-sided tumor subgroups. Since race and the proportions of patients receiving PTR and PMTR may vary simultaneously across HSAs, we also performed stratified analysis. Taking into account the influence of medical resources and other factors, the PTR rate may be higher in areas with high PMTR rate. To evaluate the treatment effect of PMTR versus other strategies (i.e., PTR, metastasectomy-only, and no surgery to the primary or metastatic sites), we applied a 2SRI stratified by race (non-Hispanic white vs. non-Hispanic black vs. others) and the proportion of patients receiving PTR (below the median vs. equal to or above the median) . In the first stage, we fitted a stratified logistic model including the HSA’s proportion of patients receiving PMTR (below median vs. equal to or above the median) to predict the receipt of PMTR versus other strategies, and calculated the residual, which was defined as the observed probability minus the predicted probability of receiving PMTR. The residual was then included in the second stage to evaluate the impact of PMTR versus other strategies on OS using a stratified Cox analysis adjusted by an individual frailty term . We further evaluated the treatment effect of PTR versus no surgery to the primary or metastatic sites using a similar 2SRI instrumental variable analysis stratified by race and the proportion of patients receiving PMTR (below median vs. equal to or above the median).
Statistical significance was set as P < 0.05 in a two-tailed test. Statistical analysis was performed using R version 3.5.1 (http://www.r-project.org). The data cleaning and cleansing and model building relied on the R packages “tidyr”, “dplyr”, “tibble”, “data.table”, “tableone”, “caret”; survival analysis and Cox analysis, with or without IV correction relied on R packages “rms”, “survival”, and “survminer”; and figure drawing for using R package “ggplot2”.
We identified 50,333 patients diagnosed with stage IV CRC between 2005 and 2015 (Additional file 1: Fig. S1).
Additional file 2: Table S1 summarizes the baseline characteristics of the patients in the different groups. Age, marital status, race, and tumor characteristics were imbalanced between PMTR and non-PMTR groups and similar results were also observed in the PTR versus non-PTR group.
Missing data for tumor characteristics (i.e., tumor size, tumor grade, T stage, and N stage) were more common in the non-PMTR group and the non-surgery group, suggesting these variables comprised a mixture of clinical/preoperative and postoperative pathologic information. Therefore, these variables were not qualified confounding factors for the survival impact of PMTR and PTR.
Multivariable Cox analysis of survival impact of PMTR/PTR in left- and right-sided CRC subgroups
The fully-adjusted Cox models are shown in Additional file 3: Table S2.
In multivariate regression analysis, after adjusting for race, age, sex, marital status, and year of diagnosis, PMTR was found to be a predictor of longer OS in both left-sided (HR = 0.60 [95% CI 0.57–0.63]; P < 0.001; Additional file 4: Fig. S2A) and right-sided (HR = 0.52 [95% CI 0.50–0.55]; P < 0.001; Additional file 4: Fig. S2B) stage IV CRC patients. Significantly better OS was seen with PTR than with no surgery in patients with left-sided and right-sided tumors (Additional file 4: Fig. S2C and S2D).
However, Durbin–Wu–Hausman tests were statistically significant (P < 0.01 for all), suggesting that unmeasured confounders may have biased the results. Therefore, we performed instrumental variable analysis.
Instrumental variable analysis of survival impact of PMTR in left- and right-sided CRC subgroups
The directed acyclic graph for the relationship between an instrumental variable (HSA), treatment (PTR or PMTR), observed and unobserved confounding factors and the outcome (OS) had shown in Additional file 5: Fig. S3.
Figure 1A and B showed how surgical options differed between high-PMTR and low-PMTR HSAs.
In both the left-sided and right-sided subgroups, the proportion of patients receiving PTR or only metastasectomy was comparable between the high-PMTR region and low-PMTR region when stratified by PTR rate. We found significant non-arbitrary correlation between the HSA and the usage rate of PMTR in both the left-sided and the right-sided subgroups, as indicated by F statistics above the standard cutoff of 10 (271.5 and 229.5, respectively).
Table 1 shows the baseline characteristics of the left-sided and right-sided tumor groups in the high-PMTR region and low-PMTR region after stratification by HSA PTR rate.
The proportion of missing data for tumor-associated variables was higher in the low-PMTR region, irrespective of PTL, because of the reasons mentioned above. For the other variables (except race), the absolute differences between the high-PMTR and low-PMTR regions were numerically small. Taken together, these findings suggest that the HSA PMTR rate satisfies the major requirements for a valid instrument. We then conducted instrumental variable analyses stratified for race and HSA PTR rate to evaluate the impact of PMTR on survival.
In the left-sided stage IV CRC group, survival was significantly better in those treated with PMTR than in those not treated with PMTR: median OS, 41 months (95% CI 32–59) versus 16 months (95% CI 15–17); HR = 0.37; 95% CI 0.24–0.58 (P < 0.001; Fig. 1C). However, in right-sided stage IV CRC, PMTR provided no significant survival benefit (HR = 0.98 [95% CI 0.65–1.47]; P = 0.910; Fig. 1D; interaction test P < 0.001).
To determine whether the different results in the two PTL groups were due to different characteristics of the respective marginal populations for instrumental variable analyses, we assessed whether an HSA with a high PMTR rate for right-sided tumors also tended to have a high PMTR rate for left-sided tumors. The HSA-level PMTR rates for left-sided and right-sided tumors showed strong positive correlation (Pearson correlation coefficient weighted by HSA-level patient number, r = 0.71; P < 0.001; Fig. 2), suggesting that the PMTR practice pattern was similar for left-sided and right-sided tumors in each HSA and that the characteristics of the marginal populations in the left- and right-sided subgroups were similar in the instrumental variable analyses.
Instrumental variable analysis of survival impact of PTR in left- and right-sided CRC subgroups
Figure 3A and B showed the differences in surgical options between the high-PTR and low-PTR regions.
The proportions of patients receiving PMTR or only metastasectomy were similar between high-PTR regions and low-PTR regions after stratification by PMTR rate. Among patients treated with PTR or no surgery, we found a robust association between HSA and use of PTR in both left-sided and right-sided CRC subgroups (F statistics of 58.2 and 58.3, respectively). Baseline covariates (except race) were also comparable between the high-PTR and low-PTR regions (Additional file 6: Table S3).
After stratification by race and HSA PMTR rate, instrumental variable analysis showed comparable survival with PTR and with no surgery among patients with left-sided CRC: median OS, 17 months (95% CI 13–21) versus 17 months (95% CI 13–21); HR = 1.11 (95% CI 0.68–1.81); P = 0.690 (Fig. 3C). Similarly, no significant difference was seen in the right-sided CRC group: median OS, 12 months (95% CI 10–15) versus 9 months (95% CI 6–15); HR = 0.85 (95% CI 0.50–1.43); P = 0.530; interaction test P = 0.466 (Fig. 3D). The correlation of PTR rate in left-sided and right-sided tumors across different HSAs was demonstrated in Additional file 7: Fig. S4.
To our knowledge, this is the first study to evaluate the causal impact of PMTR and PTR on OS in left- and right-sided stage IV CRC patients, using instrumental variable analysis to adjust for both measured and unmeasured confounders. We found significant survival benefit with PMTR only in left-sided stage IV CRC patients. It should be emphasized that this finding cannot be generalized to the entire right-sided stage IV CRC population, because results from instrumental variable analysis only pertain to the marginal patients whose choice of PMTR is directly influenced by regional practice patterns. Although it is not possible to accurately delineate the characteristics of this marginal population, it may comprise those in whom surgery cannot be performed for technical reasons although the lesions are potentially resectable. Based on the correlation between PMTR rates in left- and right-sided CRC patients (Fig. 2), we speculate that the surgery patterns in the left- and right-sided subgroups were similar across different HSAs, which suggests that the characteristics of the marginal populations in the two subgroups were similar in the instrumental variable analyses.
A possible explanation for the lack of benefit with PMTR in the right-sided subgroup is the more aggressive biological behavior of right-sided tumors. The levels of specific biomarkers, such as microsatellite instability (MSI), CpG island methylator phenotype (CIMP) level, BRAF and KRAS mutations, gradually increase from the distal to the proximal colon . A previous study has found that right-sided advanced CRC is more likely to be associated with RAS and BRAF mutations . It has been reported that BRAF V600E mutation and KRAS mutations are associated with a worse prognosis in stage IV CRC . Herein, these biological distinctions between right- and left-sided colon cancers could also translate into important differences in survival. Patients with right-sided tumors have poorer response to chemotherapy and target agents like cetuximab, and also have shorter OS after curative-intent hepatectomy for liver metastasis [19, 38]. Based on our findings, PTL should be taken into consideration when deciding on the feasibility of PMTR, with more stringent selection criteria applied to patients with right-sided tumors. One strategy could be to select for PMTR only those who have shown good response to systemic therapy. Our findings also suggest that the overall treatment outcomes in the regions that underutilize PMTR could be improved through efforts to increase use of PMTR in the marginal population with left-sided tumors.
There is considerable controversy regarding whether or not an asymptomatic primary tumor should be removed in patients with incurable stage IV CRC. Our instrumental variable analysis shows that PTR is not associate with improved survival in either left- or right-sided stage IV CRC. The marginal patients may be probably those who were asymptomatic, among whom the PTR rate would vary greatly across different regions. However, it should be mentioned that several retrospective observational studies have reported definite survival benefits with PTR in stage IV CRC [8,9,10]. The non-concordance with our results could be explained by the differences in statistical methods and the selection bias that is likely in retrospective investigations. We used HSA as instrumental variable to control for both measured and unmeasured confounders that could lead to misinterpretation of the impact of PTR and to focus on the marginal (asymptomatic) patients. Moreover, we only included patients diagnosed between 2005 and 2015, when systemic chemotherapy and targeted therapy were well developed, whereas most of the previous studies included patients treated before 2006. Some previous studies, including a recent randomized study, have reported results similar to ours, that is, PTR does not provide any survival benefit over that provided with systemic chemotherapy [11, 13, 14]; our investigation further shows that this is true for both left- and right-sided stage IV CRC patients. According to previous US population-based studies, more than half of stage IV CRC patients have received palliative PTR even in recent time [8, 12], which highlights the need to avoid the overuse of PTR. Multidisciplinary conference before PTR may help avoid unnecessary surgery and the associated delay in start of chemotherapy or targeted therapy .
There are several limitations in our study. First, our results are valid only as far as the assumptions of the instrumental variable are satisfied. We have attempted to confirm that our instrumental variable met the requirements, but there may still be some unobserved instrumental variable-related factors that were not well balanced between the groups. For instance, we cannot exclude the possibility that the variation in resection rates across HSAs is related to the quality of operation in these hospitals. However, the balanced usage of other surgical options in high- and low-PMTR/PTR areas (Figs. 1A and 3A) suggested comparable surgical quality among these areas. Second, metachronous metastasis cases were not included in our study, and thus the survival impact of PMTR and PTR remains to be explored in this subgroup. Third, instrumental variable analysis estimates only the marginal effect on the population under study, and it is difficult to define who the marginal patients are; further investigation is necessary to delineate these patients so that patient selection for PMTR can be optimized. Other limitation is the absence of mutation status and not considering this in multivariate model as these are important prognostic factors. Finally, our findings need to be further validated in other independent cohorts.
PMTR appears to provide survival benefit only for patients with left-sided stage IV CRC, while geographically increased use of PTR is not associated with improved OS regardless of PTL. Our results suggest that PTL should be taken into consideration when selecting patients for PMTR, and routine use of palliative PTR in asymptomatic patients should be avoided. Multidisciplinary conference may help in selection of the best surgical candidates. Further studies are required to identify the patients who could benefit from surgery.
Availability of data and materials
Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394–424.
Van Cutsem E, Nordlinger B, Cervantes A, ESMO Guidelines Working Group. Advanced colorectal cancer: ESMO Clinical Practice Guidelines for treatment. Ann Oncol. 2010;21(Suppl 5):v93–7.
Van Cutsem E, Nordlinger B, Adam R, Kohne CH, Pozzo C, Poston G, et al. Towards a pan-European consensus on the treatment of patients with colorectal liver metastases. Eur J Cancer. 2006;42(14):2212–21.
Prasanna T, Wong R, Price T, Shapiro J, Tie J, Wong HL, et al. Metastasectomy and BRAF mutation; an analysis of survival outcome in metastatic colorectal cancer. Curr Probl Cancer. 2021;45(1):100637.
Zaydfudim VM, McMurry TL, Harrigan AM, Friel CM, Stukenborg GJ, Bauer TW, et al. Improving treatment and survival: a population-based study of current outcomes after a hepatic resection in patients with metastatic colorectal cancer. HPB. 2015;17(11):1019–24.
Cummings LC, Payes JD, Cooper GS. Survival after hepatic resection in metastatic colorectal cancer: a population-based study. Cancer. 2007;109(4):718–26.
National Comprehensive Cancer Network. Colon Cancer (Version 1.2020). http://www.nccn.org/professionals/physician_gls/pdf/colon.pdf. Access 19 Dec 2019.
Tarantino I, Warschkow R, Worni M, Cerny T, Ulrich A, Schmied BM, et al. Prognostic relevance of palliative primary tumor removal in 37,793 metastatic colorectal cancer patients: a population-based, propensity score-adjusted trend analysis. Ann Surg. 2015;262(1):112–20.
t’ Lam-Boer J, Van der Geest LG, Verhoef C, Elferink ME, Koopman M, de Wilt JH. Palliative resection of the primary tumor is associated with improved overall survival in incurable stage IV colorectal cancer: A nationwide population-based propensity-score adjusted study in the Netherlands. Int J Cancer. 2016;139(9):2082–94.
Xu H, Xia Z, Jia X, Chen K, Li D, Dai Y, et al. Primary tumor resection is associated with improved survival in stage IV colorectal cancer: an instrumental variable analysis. Sci Rep. 2015;5:16516.
Mehta HB, Vargas GM, Adhikari D, Dimou F, Riall TS. Comparative effectiveness of chemotherapy vs resection of the primary tumour as the initial treatment in older patients with Stage IV colorectal cancer. Colorectal Dis. 2017;19(6):O210–8.
Ahmed S, Leis A, Fields A, Chandra-Kanthan S, Haider K, Alvi R, et al. Survival impact of surgical resection of primary tumor in patients with stage IV colorectal cancer: results from a large population-based cohort study. Cancer. 2014;120(5):683–91.
Alawadi Z, Phatak UR, Hu CY, Bailey CE, You YN, Kao LS, et al. Comparative effectiveness of primary tumor resection in patients with stage IV colon cancer. Cancer. 2017;123(7):1124–33.
Kanemitsu Y, Shitara K, Mizusawa J, Hamaguchi T, Shida D, Komori K, et al. A randomized phase III trial comparing primary tumor resection plus chemotherapy with chemotherapy alone in incurable stage IV colorectal cancer: JCOG1007 study (iPACS). J Clin Oncol. 2020;38(4_suppl):7.
Hu CY, Bailey CE, You YN, Skibber JM, Rodriguez-Bigas MA, Feig BW, et al. Time trend analysis of primary tumor resection for stage IV colorectal cancer: less surgery, improved survival. JAMA Surg. 2015;150(3):245–51.
Mik M, Berut M, Dziki L, Trzcinski R, Dziki A. Right- and left-sided colon cancer-clinical and pathological differences of the disease entity in one organ. Arch Med Sci. 2017;13(1):157–62.
Benedix F, Kube R, Meyer F, Schmidt U, Gastinger I, Lippert H. Comparison of 17,641 patients with right- and left-sided colon cancer: differences in epidemiology, perioperative course, histology, and survival. Dis Colon Rectum. 2010;53(1):57–64.
Yamauchi M, Morikawa T, Kuchiba A, Imamura Y, Qian ZR, Nishihara R, et al. Assessment of colorectal cancer molecular features along bowel subsites challenges the conception of distinct dichotomy of proximal versus distal colorectum. Gut. 2012;61(6):847–54.
Tejpar S, Stintzing S, Ciardiello F, Tabernero J, Van Cutsem E, Beier F, et al. Prognostic and predictive relevance of primary tumor location in patients with RAS Wild-type metastatic colorectal cancer: retrospective analyses of the CRYSTAL and FIRE-3 trials. JAMA Oncol. 2017;3(2):194–201.
Price TJ, Beeke C, Ullah S, Padbury R, Maddern G, Roder D, et al. Does the primary site of colorectal cancer impact outcomes for patients with metastatic disease? Cancer. 2015;121(6):830–5.
Rassen JA, Schneeweiss S, Glynn RJ, Mittleman MA, Brookhart MA. Instrumental variable analysis for estimation of treatment effects with dichotomous outcomes. Am J Epidemiol. 2009;169(3):273–84.
National Cancer Institute. Surveillance epidemiology and end results. http://seer.cancer.gov. Accessed 10 Oct 2019.
Goldstein DA, Zeichner SB, Bartnik CM, Neustadter E, Flowers CR. Metastatic colorectal cancer: a systematic review of the value of current therapies. Clin Colorectal Cancer. 2016;15(1):1–6.
Arnold D, Lueza B, Douillard JY, Peeters M, Lenz HJ, Venook A, et al. Prognostic and predictive value of primary tumour side in patients with RAS wild-type metastatic colorectal cancer treated with chemotherapy and EGFR directed antibodies in six randomized trials. Ann Oncol. 2017;28(8):1713–29.
Hernan MA. A definition of causal effect for epidemiological research. J Epidemiol Community Health. 2004;58(4):265–71.
Vanderweele TJ, Arah OA. Bias formulas for sensitivity analysis of unmeasured confounding for general outcomes, treatments, and confounders. Epidemiology. 2011;22(1):42–52.
VanderWeele TJ, Mumford SL, Schisterman EF. Conditioning on intermediates in perinatal epidemiology. Epidemiology. 2012;23(1):1–9.
Lash TL, Fox MP, Fink AK. Applying quantitative bias analysis to epidemiologic data. Berlin: Springer; 2009.
Terza JV, Bradford WD, Dismuke CE. The use of linear instrumental variables methods in health services research and health economics: a cautionary note. Health Serv Res. 2008;43(3):1102–20.
Terza JV, Basu A, Rathouz PJ. Two-stage residual inclusion estimation: addressing endogeneity in health econometric modeling. J Health Econ. 2008;27(3):531–43.
Sheetz KH, Norton EC, Regenbogen SE, Dimick JB. An instrumental variable analysis comparing medicare expenditures for laparoscopic vs open colectomy. JAMA Surg. 2017;152(10):921.
Raoof M, Haye S, Ituarte PHG, Fong Y. Liver resection improves survival in colorectal cancer patients: causal-effects from population-level instrumental variable analysis. Ann Surg. 2019;270(4):692–700.
Newhouse JP, McClellan M. Econometrics in outcomes research: the use of instrumental variables. Annu Rev Public Health. 1998;19(1):17–34.
Lu-Yao GL, Albertsen PC, Moore DF, Shih W, Lin Y, DiPaola RS, et al. Fifteen-year survival outcomes following primary androgen-deprivation therapy for localized prostate cancer. JAMA Intern Med. 2014;174(9):1460–7.
Martinez-Camblor P, Mackenzie T, Staiger DO, Goodney PP, O’Malley AJ. Adjusting for bias introduced by instrumental variable estimation in the Cox proportional hazards model. Biostatistics. 2019;20(1):80–96.
Missiaglia E, Jacobs B, D’Ario G, Di Narzo AF, Soneson C, Budinska E, et al. Distal and proximal colon cancers differ in terms of molecular, pathological, and clinical features. Ann Oncol. 2014;25(10):1995–2001.
Passiglia F, Bronte G, Bazan V, Galvano A, Vincenzi B, Russo A. Can KRAS and BRAF mutations limit the benefit of liver resection in metastatic colorectal cancer patients? A systematic review and meta-analysis. Crit Rev Oncol Hematol. 2016;99:150–7.
Sasaki K, Andreatos N, Margonis GA, He J, Weiss M, Johnston F, et al. The prognostic implications of primary colorectal tumor location on recurrence and overall survival in patients undergoing resection for colorectal liver metastasis. J Surg Oncol. 2016;114(7):803–9.
Scheer MG, Sloots CE, van der Wilt GJ, Ruers TJ. Management of patients with asymptomatic colorectal cancer and synchronous irresectable metastases. Ann Oncol. 2008;19(11):1829–35.
The authors acknowledge the efforts of the Surveillance, Epidemiology, and End Results (SEER) Program tumor registries in providing high quality open resources for researchers.
This work was supported by: National Natural Science Foundation of China (81930065); Natural Science Foundation of Guangdong Province (2014A030312015); Science and Technology Program of Guangdong (2019B020227002); Science and Technology Program of Guangzhou (201904020046, 201803040019, 201704020228). Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Ethics approval and consent to participate
This study was approved by the Ethics Committee (B2022-138-MS01) of the Sun Yat-sen University Cancer Center (Guangzhou, China) and complied with the Helsinki Declaration. SEER is public-use data, no institutional review was sought and informed consent was waived.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Study Flow Chart.
Patient Characteristics in the Different Groups. Abbreviations: PMTR, primary tumor resection plus metastasectomy; PTR, primary tumor resection.
Fully-adjusted Cox models. Abbreviations: HR, hazard ratio; CI, Confidence intervals.
Multivariable Cox Analysis Evaluating the Impact of PMTR and PTR on Overall Survival. Abbreviations: CRC, colorectal cancer; HR, hazard ratio; OS, overall survival; PMTR, primary tumor resection plus metastasectomy; PTR, primary tumor resection.
The directed acyclic graph.
Comparison of Baseline Characteristics Between the High-PTR Region and Low-PTR Region, Stratified by HSA PMTR Rate. Abbreviations: HSA, health service area; PMTR, primary tumor and metastasis resection; PTR, primary tumor resection.
Correlation of PTR Rate in Left-sided and Right-sided Tumors across Different HSAs. Abbreviations: PTR, primary tumor resection.
About this article
Cite this article
Yao, YC., Chen, JQ., Yin, L. et al. Primary tumor resection with or without metastasectomy for left- and right-sided stage IV colorectal cancer: an instrumental variable analysis. BMC Gastroenterol 22, 114 (2022). https://doi.org/10.1186/s12876-022-02184-2