Scribd Logo
Upload

Welcome to Scribd!

  • Circumferential Resection Margin As A Prognostic Factor After Rectal Cancer Surgery A Large Population - Based Retrospective Study

    Uploaded by

    Keam Pisethsapheareach
    13 views9 pages

    Original Title

    Circumferential Resection Margin as a Prognostic Factor After Rectal Cancer Surgery a Large Population- Based Retrospective Study

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    13 views9 pages

    Circumferential Resection Margin As A Prognostic Factor After Rectal Cancer Surgery A Large Population - Based Retrospective Study

    Uploaded by

    Keam Pisethsapheareach

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    of 9
    ORIGINAL RESEARCH WILEY CancerMedicine Circumferential resection margin as a prognostic factor after rectal cancer surgery: A large population-based retrospective study QiLiu’? | Dakui Luo? | "Deparment of Colorectal Surgery, Fudan Universi Shanghai CneeeCeater, Shanghai, China ‘Deparement of Oncology, Shang Medical College, Fudan University, Shanghai, China Correspondence: Xiniang Li and (inggv0 Li, Department of Colorectal Surgery, Fadan University Shanghai Cancer Center, #270 Dongan Rox, Xubss Dstret, Sanghi, 200082, China (1149)xx@sina, Funding information This researc was supported by the [National Science Foundation of China (No, 181702853 and 81772598) and Shanghat Masicipal Natal Seience Foundation (7781405100), The funders had no le in the study design, data eolletion and snalysis, decision to pubis, or preparation of the manuscript. 1 | INTRODUCTION (Circumferential resection margin (CRM) is the closest distance ‘between the radial resection margin and the tumor issue by either Sanjun Cai!?@ | Qingguo Li’? | Xinxiang Li’? Abstract ‘Aim: This study aimed to investigate circumferential resection margin (CRM) as a ‘prognostic factor for long-term oncologic survival after rectal cancer surgery. Methods: Patients diagnosed with malignant rectal cancer between 1 January 2010 and 31 December 2014, from the Surveillance, Epidemiology, and End Results (SEER) program were identified for this study. The patients were divided into five CRM groups to compare the baseline characteristics and assess cancer-specific sur. vival (CSS): 0-1 mm, 1.1-2.0mm, 2.1-5.0mm, $.1-10.0 mm, and >10 mm. The main endpoint was CSS, Results: Circumferential resection margin <1 mm was independently associated with 99% increased risk of cancer-specific mortality in rectal cancer [hazard ratio (HR) = 1.990, 95% confidence interval (CI) = 1.613-2.454, P< 0.001, using CRM (1.1.2.0 mm) as a reference]. CRM (5.1-10.0 mm) was independently associated ‘with 29.2% decreased risk of cancer-specific mortality [HR = 0.708, 95% CI = 0.525- 0.954, P = 0.152, using group (2.1-5.0 mm) as reference]. CRM > (QLiw and Dak Lvo conbaed equally to he work, work i prope ee 2 L Wwitey—CancerMedicine While several studies showed that CRM should not be used as a prognostic factor in rectal cancer,** other studies demonstrated the importance of CRM as an independent prognostic factor of local recurrence and long-term sur- vival," ineluding the first repost by Quirke etal" suggesting that CRM might be a strong predictor of long-term oncologic outcomes. Avcording to the European Society for Medical Oncology (ESMO) Clinical Practice Guidelines for rectal cancer, CRM is defined as involved if itis <1 mm from the tumor free margin, leading to an inereased risk of local recur- rence, distant metastases, and poorer survival Many studies considered CRM as positive when it was <1 mm (R1) and associated with obviously poor prognosis as compared to CRM >1 mm (RO), whieh was in accordance with the ESMO guidelines.“*-"” The criterion to define a positive CRM remains unclear: However, some researchers believed that CRM within 2 mm was associated with a nega- tive prognosis"? In addition, Kelly et al! argued that a CRM clearance >5mm should be achieved to optimize curative tweatment, Recently, Beauftére etal found that the prognosis after rectal cancer surgery was worse with a CRM <0.4 mm. Given that the aforementioned studies had a relatively small sample size, the Surveillance, Epidemiology, and End Results (SEER) program conducted a large population-based study to analyze the prognostic ability of CRM distance in rectal cancer. 2 | PATIENTS AND METHODS 2.1 | CRM in SEER database ‘The SEER database is an authoritative source of information fon the most recent cancer incidence, mortality, prevalence, and lifetime risk statistics in the United States. It is a com- prehensive sousce of population-based information including all newly diagnosed cancer cases among people residing in SEER-participating areas and covering approximately 28% of the US population. ‘The CRMs in SEER database are, expressed as the nearest tenth in millimeters (mm), the distance between the leading edge ofthe tumor and the nearest edge of surgically dissected. margin, a8 recorded in the pathology report according 10 The American Joint Committee on Cancer (AJC) Seventh. Edition Cancer Staging Manual: the CRM is the surgically dissected nonperitonealized surface of the specimen, 22 | Using the SEER-Stat software (SEER*Stat 8.3.4), patients diagnosed with malignant sectal cancer between I January 2010 and 31 December 2014, from the SEER Program of the National Cancer Institute were identified (Figure 1) Among these patients, patients with known CRM (eg, CS Study design Site-Specific Factor 6, 2004, varying by Schema 000, 032, and 981) were included in the analysis. Patients with rectal cancer whose CS Site-Specific Factor 6 was 996 (means ‘>5 mm’; cannot be grouped in this study), with unknown seventh AIC stage or unknown race record, were excluded. Patients were then divided into five CRM groups: 0-1 mm, 1.1-2.0 mm, 2.1-5.0 mm, $.1-10.0 mm, and >10 mm to com- pare the baseline characteristics and assess cancer-specific survival (CSS). Next, patients whose CS Site-Specific Factor 6 was 000 (CRM 10 mm. 23° | Several Cox proportional hazards models were built to iden- tify independent prognostic variables at a median survival time of 22 months (range 0-59 months). All hazard ratios (AR) were shown with 95% confidence intervals (CI). A mul- tivariate survival analysis was performed using a Cox propor- tional hazards model, including all variables associated with a P value <0.2 in univariate analysis. Variables including AIC stage, tumor size, age at diagnosis, race, gender, year of diagnosis, and grade were included in the Cox multivariate survival analysis. The TNM staging used in the present study was the seventh edition of the AJCC cancer staging system, the newest TNM staging that could be obtained from the SEER database, The primary outcome of interest was CSS. ‘The Kaplan-Meier survival curves were used to evaluate the prognostic prediction of different factors. The log-rank tests were used to assess statistical significance, All tests were two sided, and P values <0.05 were considered statistically significant. Statistical analysis was performed using Statistic Package for Social Science (SPSS) version 22 (SPSS Inc... IL, USA). Statistical analyses 3 | RESULTS 3.1 | Patient characteristics of the overall cohort ‘A total of 10 181 patients with rectal cancer after surgery ‘were identified from the SEER database. The baseline de- ‘mographic characteristics of the patients are summatized in Table 1. A total of 4232 (41.6%) patients whose CRM was between 0 and | mm were included in the analyses. The over- all cohort showed that higher AIC stages (P < 0,001), larger tumor size (P < 0,001), black people (P < 0.001), earlier year of diagnosis (P < 0.001), and higher grades (P < 0,001) were associated with a CRM between 0 and 1 mm. Differences in other characteristics were not significant. CancerMedicine WILEY? Rectal cancer patent t SEER 18 regities Between Janay 1, 020, and Decamber 31, 2018 sclsion: w= ss6ss) (fog, CS St specif Factor 5 3) (9 #10662) (Ste apc Factor 2008 Varying by Sehoms (rans “eater than Sm cannot be ‘grouped ints say n= 295) “Te TNM stage s unknown or stage = 0(0= 150) ‘ace record eunkown (N= 36). Population of tho frst analy ‘n=30083) (5 Site specie Factor 6 2008 Varying by sehema = 000 ‘vcude (CRM © ren and tnowed FIGURE 1 Flow diagram of patient, population selection from the SEER database 3.2. | R1CRM was strongly associated with poor survival in rectal cancer The median follow-up duration for the overall cohort was 22 months (range 0-59 months). At the end of the follow-up, 1262 (12.4%) patients died of rectal cancer. ‘A multivariate analysis was conducted to identify the variables independently associated with CSS in the overall cohort, The results of multivariate analyses by Cox regres- sion ate detailed in Table 2. R1 CRM was found to be inde- pendently associated with CSS of 10 181 patients with rectal cancer and bad a 99.0% increased risk of cancer-specific mortality [HR = 1.990, 95% CT =1.613-2.454, P< 0.001, using group (1.1-2.0 mm) asa reference]. In addition, Table 2 shows that lower AJCC stages, younger age, and lower grades were independent protective factors. Kaplan-Meier CSS curves were used to analyze the prog- nosis of different CRMs (Figure 2). Group (0-1.0 mm) was associated with poorer CSS (90.0% for 1-year CSS and 73.8% for 3-year CSS). The 3-year CSS of group (.1-2.0 mm), group (21-5.0mm), group (5.1-10.0 mm), and group Population of father anti ona) ($10.0 mm) was 88.2%, 873%, 91.4%, and 90.8%, respec- tively, However, the differences between group (11-20 mm) and group (2.15.0 mm), group (1. 1-2.0 ram) and group (5.1- 10.0 mm), and group (6.1-10.0 mm) and group (>10.0 mmm) were not statistically significant, The Cox multivariate CSS analysis also showed no significant difference between group (21-50 mm) and group (11-20 mm) [HR =0,905, 95% €C1=0,700-1,169, P= 0.483, using group (2.1-5.0 mm) as reference] (Table $1). However, group (6.1-10.0 mm) had more favorable prognosis as compared to group (21-5.0 mm) (HR = 0.708, 95% CI = 0.525.0.954, P = 0.152, using group @.1-5 0 mm) as reference. Kaplan-Meier OS curves were also used to analyze the prognosis of different CRMs (Figure 3). Group (0-1.0 mm) was associated with poorer OS (84.9% for L-year OS and 62.4% for 3-year OS). The 3-year OS of group (1.1-2.0 mm). group (2.1-5.0mm), group (5.1-10.0mm), and group ($100 mm) was 78.5%, 79.0%, 81.3%, and 83.8%, respec tively. However, the differences between group (1.1-2.0 mm) and group (215.0 mn), group (I.1-2.0 mm) and group (5.1- 10.0 mm), group (21-5.0 mm) and group (51-100 mm), and +L witey—Cancer Medicine ae TABLE 1 Comparison ofbseline characterises by various cizcunferental resection margins Distance to circumferential resection margin Variable 010mm 120mm 21-50mm S1100mm ——>10.0mm Pale AICC stage emonem 2141138) 355.1885) 1668.85) s31082%) <0001 1071 693%) 3201188) 4281578) 274 001%) 630 03.1%) Stage I rman ——_4x212%) 671. 056%) 426.9%) 957 22%) Stage 1V 769 60.8%) 1106.75) 127 000%) 84.60%) 174 03.88) Tumor size sem 2x79079%) —TOH(LL2) 10390665) 206%) 1549@47R) —Sem 1499.477%) 315 400%) 44404.18) 283 00%) 01 09.1%) Unown 354462% 107 (180%) s8.1128%) 65 (83%) 142 085%) {Age a diagnosis Q) 60 1903 414%) 5041108) 700 1528) 43304%) 1061 @3.1) 0757 > 229178) 21s) 581 1585) SI7@3% aL @2.%) Race White sss 09%) 93212%) B18 0588) T79(@4%) 1876 26%) <0.001 Black 408 61.3%) 34 (106%) 104 03.18) 54 68%) 145 182%) otter 436097) 110,100%) 164 04.9%) 117 007%) amanrs) Gender Mate 2583 (20%) o88 L4) 907 15.0%) se2@3%) —1384024%) 0287 Female 1689 (409%) 438 106%) 674.0638) 388.3%) 938 @2.7%) ‘Year of dagnois 2010 s95 (485%) 2251228) 290(05.7%) 157(85%) 280(1528) <0001 201 s7429%) 2294138) siTase%) msm 48 218%) 2012 s3413% 216 (402% 525 154%) 20808) 491 232%) 2013 s03088% 220 (406%) si7ass%) 213 40.3%) S16 249%) 201s wa1G72% 236 1.1% 32 05.7%) 200.9.4%) 562 26%) Grade Grade 279 (63.7%) 7501188) 190725) 61065) 134778) <0.001 Grade 283581%) —8S9C15% 12170638) 730(08%) 1806 043%) Grode It 725 65.5%) 116(89%) 165 (27%) 38 (67%) 210 (16.1%) Grade IV 170 60.1%) 25 (68%) 105%) 18 64%) 430528) Ualaown 23 (34%) 512008) e228) 53 104%) 120 2368) group (5.1-10.0 mm) and group (>10.0 mm) were not statis- tically significant 33 | Patients whose CRM was not known to be <0.4 mm. Further analysis of RI CRM were excluded, Hence, 7811 patients with rectal cancer were identified for further analysis of RI CRM. R1 CRM. was further divided into group (<0.4mm) and group (0.5.1.0 mm) ‘A multivariate analysis was conducted to identify whether the CSS was different between group (<0.4 mm) and group (0.5-1.0 mmm) in this target population, The results of multivariate analyses by Cox regression are detailed in ‘Table 3. The difference between group (<0.4 mm, n= 741) and group (0.5-1.0 mm, n= $76) was HR = 0.834, 95% CI 0.645-1.078, using group (<0.4 mm) as reference, but it was not statistically significant (log-rank test, P = 0.166) 4 | DISCUSSION While CRM is widely accepted as a strong independent prognostic factor of long-term oncologic survival, the eri- terion used to define a positive CRM remains controver- sial."® Therefore, this retrospective study was conducted exw WiLeylt TABLE 2 Mulivariate Cox regression analyses of CSS to study CRM 100mm 825 (0687-107) 0.432.187 AVC stage Stage Stage 2.426 (1.769.327) 0.161 <0.001 Stage I 4.159 (.538-6400) 0.151 <0.001 Stage IV 15909 o1s¢ <0.001 (11773-21499) ‘Tumorsize Unknown Sem Li29 (0908-1411) 0.114 0284 Ageat <60 >60 L777 (1584-1993) 0058 <0.001 diagnosis) Year of 2010 201 1124 (0973-1299) 0074 0.111 Aiagnosis 2012 0884 (0747-1016) 0086 0.152 2013, 1014(0832-1237) 0.010887 21 9830 (0.611-1128) 0156 0.234 Gnde Gadel Grade 1286 (0942-1757) 0159 0113 Grede 1 2.195 (1.584308) 0.166 <0.001 Grade TV 27931.914-4078 0.193 <0.001 Vaknown 1.439(0963-2.150) 02050076 CRM 0-10 mm 11-20mm 21-50 mm 53-100 mm 11-20mm _ <0001 21-50mm <0001 0494 51-100mm <0901 0092 0.030 100mm <0001 0014 ogoz 0.766 20 100 Eo Eo - RW 3 ae oto mm 0mm S = Som 3 == 54400 mm e Two sided log rank st P values = Homm é ° 0 20 0 40 50 eo ‘Survival time (mo) Number at risk Group: 0-1.0 mm ‘4092 2988 225 1304 73 333 ° Group: 1-20 mm ‘098 229 629 422 24 7 ° Group: 2.45.0 mm 1546 “7 864 605 334 134 ° Group: 5:1-10.0 mm 925 m8 523 at 198 as ° Group: >10.mm 2244 1668 1194 798 433 10 ° FIGURE 2 Kaplan-Meier cancer-specific survival cave according to circumferential resection margin (CRM) 100. 80. 60. 40. crm po] dd-29.mm 21-50mm 10.0 mm ° Number at risk Group: 0-1.0 mm ‘4092 Group: 1.1-2.0 mm {ose Group: 2.1-5.0 mm ‘sas 4-10. mm 925 “Two sided log rank test P values 0-10 mm 11-20mm 21-50mm 51-100 mm <0.001, <0.001 51-100 mm <0.001, 0.001 10 2088, 229 a7 ne 1668 one 010 <0001 20 2128 «29 64 523 1194 0.63 <0001 0.081 30 40 Survival time (mmo) 1304 783 aa 258 0s a4 at 198 798 433 50 333 “7 134 as 470 FIGURE 3 Kaplan-Meier overall suvival curve according to citcunsferential resection margin (CRM) Variable Reference Characteristic cRM 100 mm ASC stage Stage Stage IE Stage HT Stage IV Tumorsize Unknown Sem >5em Ageat <60 >60) diagnosis Q) Grade GradeT Grade It Grade 1 Grade TV Vaknown Cancer-specifie survival HR (95%CD) (0.834 (0645-1078) 0.704 (0536-0924) 0.773 (0.602-0.992) 0543 (0397-0741) 0.577 (0448-0749) 2.041 (1.388-3.001) 4.361 (.060-6216) 15773 (40 949-22.722) ‘0.941 (0.685-1.292) 1.283 (0927-1.775) 1.962 1.674-2.298) 1.125 (0.738-1.714) 2.077 1.331-8241) 2.899 (1.685-4 987) 1.027 0.584-1.805) to assess the influence of CRM on prognosis after rec- tal cancer surgery, The present study included more than, 10 000 patients with rectal cancer, which greatly exceeded SE. oust 0139 oaz7 0438 0129 0197 oust 0.186 0162 0.166 081 oats 0227 oan 0288 Pvalue 0.166 oor 083 <0.001, <0.001, <0.001 <0.001 -<0.001 0707 033 -<0.001 0583 00 <0.001, 0.926 TABLE 3 $ LWwitEy—CancerMedicine cRW 040mm = 142.0mm — 215.0mm 5:1-10.0 mm >40mm Molivaiate Cox regression analyses of CSS to study CRM <0.4 mm the number of cases in previous studies and hence the re- sults of this study are persuasive and depict real-world R1 CRM was found to be an independent factor for poor prognosis and had 99.0% increased risk of cancer-specific ‘mortality as compared (© group (.1-2.0 mm). This is con- sistent with most studies on the prognostic prediction of crm However, some researchers argued that CRM <2 mm was associated with a negative prognosis.""” A CRM clearance of >5 mm and 0.4 mm was proposed by Kelly etal! in 2009 and Beaufréve et a” in 2017, respectively. Relevant analyses were also conducted in this study to assess previous study results. The Cox multivariate analysis showed that the dif ferences in CSS between group (I.1-2.0 mm) and group (2.1.5.0 mm) were not statistically significant. Yet, group (5.1-10.0 mm) had 29.2% decreased cancer-specific mor tality as compared to group (2.1-5,0 mm). After excluding the patients whose CRM was <0.4 mm or unknown, the Cox multivariate analysis showed no statistically significant dif- ference between group (<04 mm) and group (0.5-1 mm). Given the large sample size in the present study, it was be- lieved that 2 and 0.4 mm were not optimal cutoff values, in partial agreement with Kelly etal. In the study by Kelly etal, the multivariate analysis showed 32.4% increased cancer- specific mortality in group (>1 and <5 mm) as compared to group (>5 and <10 mm), which was similar to the result of the present study. However, there was no obvious difference in CSS between group (0-1.0 mm) and group (1.1-2.0 mm) in the present study. ‘The treatment modalities have dramatically changed in the recent years. The introduction of newer surgical tech- niques (TME and laparoscopy) and neoadjuvant chemora- iotherapy have reduced the incidence of positive CRMs.” Well-performed TMEs with a resection margin on the me- sorectal plane showed <10% of margin positivity. The European Organization for Research and Treatment of Cancer trial showed that neoadjuvant radiochemotherapy had 9% decreased margin positivity as compared to short-course radiotherapy.”* MRT is the most accurate method for preoperative diag- nosis of rectal cancer and can detect tumor invasion.”” The results of the present study suggested that neoadjuvant radio- chemotherapy should be considered if the distance of tumor and the mesorectal fascia is predicted to be <1 mm by pre- operative MRI in rectal cancer. While some recent studies have reported that postoperative treatment did not improve outcomes in this situation,” we hypothesize that CRM could guide postoperative treatment in combination with preoperative MRI assessment and neoadjuvant chemother apy.’ Also, the prognosis is typically better when the dis- tance of the (umor is larger from the radial resection margin, ‘Therefore, surgeons should try to maximize the CRM and at Teast I mm of CRM should be reached. Given the 32.4% in- creased cancer-specific mortality in group (>1, <5 mm) as compared to group (>5, <10 mm), 5mm of CRM should CancerMedicine WILEY also be considered. Whether patients with the distance be- tween tumor and the mesorectal fascia predicted as less than ‘Smm by preoperative MRI need neoadjuvant chemoradio- therapy or with CRM <5 mm need more intensive postop- erative attention should depend on the individual situations The present study also found that higher AJCC stages, larger tumor size, black people, earlier year of diagnosis, and higher grades were associated with a CRM between 0 and 11mm, resulting in poor prognosis. AJC stage, tumor size, and tumor grade are known prognostic factors in rectal can- cer, adding to the evidence that CRM is strongly associated ‘with the prognosis in rectal cancer. The incidence of RI ‘CRM is reducing every year due to the improvements in treat ‘ment. Black people are more likely (o achieve RI CRM and should receive adequate attention. This is attributed to the fi- ‘nancial conditions and biological differences between races, This study was the largest tll date and included more than 10 000 patients for the analyses of the prognostic prediction of CRM and was the first to simultaneously analyze postop- erative RI CRM and RO CRM in depth. ‘This study had several limilations, First, the SEER da- tabase lacked the data on local recurrence, which is an im- portant factor that influences the survival of rectal cancer, However, patients with an R1 CRM often die from metastatic disease before local recurrence.”"* In addition, definitions of local recurrence were different in previous studies, mak- ing it difficult to examine the prognosis of different CRMs. ‘Therefore, CSS is thought to be a more robust endpoint to assess the prognostic prediction of CRM.” Second, the lack of factors influencing the treatment might have affected the results to some extent, However, the large sample size could offset this influence, Further, the lack of preoperative treat- ‘ment had minimal effect since CRM in the present study was ‘measured post-operation and would have improved after pre- ‘operative treatment. The longest follow-up duration was only 59 months, not exceeding 5 years. Besides, the present analy sis was solely based on retrospective data. Hence, prospective

    You might also like