- Email: [email protected]
Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands
Available online at www.sciencedirect.com
ScienceDirect EJSO xx (2015) 1e8
www.ejso.com
Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands okkerink a, M.J. Aarts b, V.E. Lemmens b, J. Homan a,*, G.M. B€ G. van Lijnschoten c, H.J. Rutten d,e, J.H. Wijsman f, I.D. Nagtegaal g, J.H.W. de Wilt a a
Department of Surgery, Radboud University Medical Center Nijmegen, Nijmegen, The Netherlands b Netherlands Comprehensive Cancer Organization, Eindhoven, The Netherlands c Department of Pathology, PAMM, Eindhoven, The Netherlands d Department of Surgery, Catharina Hospital, Eindhoven, The Netherlands e Department of Surgery, Maastricht University Medical Center, Maastricht, The Netherlands f Department of Surgery, Amphia Hospital, Breda, The Netherlands g Department of Pathology, Radboud University Medical Center, Nijmegen, The Netherlands Accepted 9 July 2015 Available online - - -
Abstract Background: Since the introduction of total mesorectal surgery the outcome of rectal cancer patients has improved significantly. Involvement of the circumferential resection margin (CRM) is an important predictor of increased local recurrence, distant metastases and decreased overall survival. Abdomino perineal excision (APE) is associated with increased risk of CRM involvement. Aim of this study was to analyze reporting of CRM and to identify predictive factors for CRM involvement. Methods: A population-based dataset was used selecting 2153 patients diagnosed between 2008 and 2013 with primary rectal cancer undergoing surgery. Variation in CRM reporting was assessed and predictive factors for CRM involvement were calculated and used in multivariate analyses. Results: Large variation in CRM reporting was found between pathology departments, with missing cases varying from 6% to 30%. CRM reporting increased from 77% in 2008 to 90% in 2012 ( p < 0.001). CRM involvement significantly decreased from 12% to 6% over the years ( p < 0.001). In multivariate analysis type of operation, low anterior resection or APE, did not influence the risk of CRM involvement. Clinical T4stage [odds ratio (OR) ¼ 3.51; 95% confidence interval (CI) ¼ 1.85e6.65) was associated with increased risk of CRM involvement, whereas neoadjuvant treatment (5 5 gray radiotherapy [OR 0.39; CI 0.25e0.62] or chemoradiation therapy [OR 0.30; CI 0.17e0.53]) were associated with significant decreased risk of CRM involvement. Conclusion: Although significant improvements are made during the last years there still is variation in reporting of CRM involvement in the Southern Netherlands. In multivariate analysis APE was no longer associated with increased risk of CRM involvement. Ó 2015 Elsevier Ltd. All rights reserved.
Keywords: Rectal cancer; Circumferential resection margin (CRM); Involvement; Abdomino perineal excision (APE); Low anterior resection (LAR)
Introduction The circumferential resection margin (CRM) is an important prognostic factor in rectal cancer care and a * Corresponding author. Department of Surgery, Radboud University Medical Center Nijmegen, Geert Grooteplein-Zuid 10, 6525 GA, Nijmegen, The Netherlands. Tel.: þ31 0 24 361 1111. E-mail address: [email protected] (J. Homan).
predictor of local recurrence, distant metastases and overall survival.1,2 The CRM was first described in 1986 by Quirke3 and is part of a standardized histopathological protocol. Transverse sectioning of the excision specimen is one of the key procedures in this protocol, which is now recommended in almost every national rectal cancer guideline in Europe.
http://dx.doi.org/10.1016/j.ejso.2015.07.015 0748-7983/Ó 2015 Elsevier Ltd. All rights reserved. Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
2
J. Homan et al. / EJSO xx (2015) 1e8
Since the introduction of the CRM, the prognosis of patients with rectal cancer has dramatically improved. The introduction of total mesorectal excision (TME) and extralevator techniques for lower rectal tumors have led to a decrease in CRM involvement, and consequently to less local recurrence and improved metastasis free and overall survival.4 Besides improved surgical techniques, the introduction of standard magnetic resonance imaging (MRI) has enabled clinicians to properly select patients with locally advanced disease and treat them with preoperatively with (chemo)radiation therapy.5,6 Also, short course radiotherapy has proved to decrease local recurrence rates,7,8 similar to long course chemoradiation therapy for locally advanced tumors.9 Besides prognostic information, the CRM provides the surgeon of essential feedback on the quality of surgery.3 Determination of predictive factors for CRM involvement is essential to provide the best patient care. According to the literature more extensive tumors (i.e. T4 and N2 stage tumors) are related to higher CRM involvement rates.10e12 The results of the Dutch TME trial illustrated a significant difference in CRM involvement between low anterior resection (LAR) and abdomino perineal excision (APE), respectively 14% and 29% ( p < 0.001).2 On the other hand, the introduction of new surgical techniques, such as extralevator APE (ELAPE), resulted in less CRM involvement and better oncological outcome in patients with distal rectal cancer.13,14 Population-based data on CRM reporting and involvement are rare in the available literature.15 The aim of the current study was to evaluate variation in CRM reporting within different pathology departments and CRM involvement in different hospitals. Moreover, prognostic factors for CRM involvement were identified using the population-based database of the Eindhoven Cancer Registry (ECR). Patients and methods All patients diagnosed with primary rectal cancer (stage IeIII) between January 2008 and January 2013 were selected using population-based data from the Eindhoven Cancer Registry (ECR) which is part of the Netherlands Cancer Registry and maintained by the Netherlands Comprehensive Cancer Organization (IKNL). The ECR collects data on all newly diagnosed patients with rectal cancer in the Southern Netherlands. The ECR covers an area with ten community hospitals and six pathology departments. Due to intensive training of the registrars and computerized consistency checks, the quality of the retrospective data is high.16 Consent was obtained from all patients in accordance with the local and international legislation (declaration of Helsinki). Patients who were diagnosed in another region, but operated in one of the ECR hospitals were not included, as well as patients who did not undergo an operation. Other exclusion criteria were stage IV rectal cancer and patients who underwent a local excision of their rectal tumor.
The outcome of interest was CRM reporting and CRM involvement. Rectal cancer was defined as a carcinoma of the rectum within 15 cm of the anal verge. In the present study a negative or free CRM was defined as more than 1 mm, or when “free” was reported in the pathology report. A positive resection margin was defined as a margin of 1 mm or less, or when “positive” was reported in the pathology report. The CRM reporting was assessed by the year of diagnosis as well as for pathological laboratories of the ECR area. Other variables that were collected for analyses were: comorbidity according to the Charlsons comorbidity index, tumor characteristics, neoadjuvant therapy, type of surgery, and localization of the tumor (topography). Neoadjuvant therapy was divided in four groups; no neoadjuvant therapy, short term 5 5 gray (Gy) radiotherapy (RT), chemoradiation therapy (CRT), and only chemotherapy. The type of surgery was divided into four groups: the LAR, APE, Hartmann operations and other surgery (which included exenterations and proctocolectomies). The localization of the tumor was categorized into three groups (based on MRI if available): the lower-rectum (distance from anus 0e49 mm), the midrectum (distance from anus 50e99 mm) and the upperrectum (distance from anus 100 mm). For describing clinical tumor characteristics the following abbreviations are used; clinical tumor stage (cT), clinical node stage (cN). In the analysis to determine predictive factors for CRM involvement all patients with missing pathology reports were excluded. In the multivariate logistic regression analyses for CRM involvement adjustments were made for sex, age category (18e49, 50e64, or age more than 65 years old), cT, cN, the use of neoadjuvant thereapy (no therapy, 5 5 Gy RT, or CRT), distance of the tumor to the anal verge (lower, mid, and upper rectum), type of surgery (LAR, APE, Hartmann, or any other type of surgery), and surgical approach (open, laparoscopy, or conversion to open surgery). The derived odds ratios (ORs) and the 95% confidence intervals (CIs) are presented. Furthermore, it was evaluated whether the number of operations on rectal cancer per hospital per year influenced CRM involvement. To do so, hospitals were divided into three groups: hospitals in which 20 operations, 21 to 40 operations or >40 operations per year were performed. Because of centralization of cT4-stage tumors to specialized centers they were excluded from this analysis to avoid negative selection bias. All statistical analyses were performed using SPSS version 21.0 for Windows (Inc., Chicago, Illinois, USA). P-values were two-sided and values < 0.05 were considered significant. Results Patients and characteristics A total of 3348 patients were diagnosed with rectal cancer during the study period. In total 71% of the patients
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
J. Homan et al. / EJSO xx (2015) 1e8
3
were surgically treated with (beyond) TME surgery, and after exclusion of 224 stage IV patients a total of 2153 patients were included. In 1833 (85%) patients the CRM was reported in the pathological report. Mean age at time of surgery was 66 11 years and 63% of the patients were male. Table 1 gives an overview of different patient and tumor characteristics and CRM involvement rates. In total 89% of the patients received neoadjuvant therapy; 52% received 5 5 Gy RT and 37% CRT. In early tumors (cT1 and cT2) CRT was only used in 16% of the patients, whereas the use of CRT was 76% for cT4 tumors (Table 2).
CRM reporting
Table 1 The baseline characteristics of patients with rectal cancer registered in the ECR area. All patients with an unreported CRM in the pathology report were excluded from this analysis.
CRM involvement rate decreased significantly over the years, from 12% in 2008 to 6% in 2012 ( p < 0.001). In total 87% of the operations were performed within the ECR area, in the other 13% the hospital was unknown or not affiliated to the ECR area. The mean number of operations for rectal cancer ranged between the hospitals from 5 to 68 operations per year. There was one hospital in which less than 20 operations for rectal cancer per year were performed. After exclusion of the cT4 tumors there was a non-significant trend towards an increased CRM involvement of 13% in that hospital compared to 6% CRM involvement in the hospital in which more than 40 operations per year were performed ( p ¼ 0.069). The multivariate logistic regression analysis for CRM involvement is presented in Table 3. A significantly decreased risk of CRM involvement was observed for patients who received neoadjuvant CRT (OR ¼ 0.30; 95% CI ¼ 0.17e0.53) or 5 5 Gy RT (OR ¼ 0.39; 95% CI ¼ 0.25e0.62) compared to patients who did not receive neoadjuvant therapy. Clinical T4 stage was associated with an increased risk on CRM involvement (OR ¼ 3.51; 95% CI ¼ 1.85e6.65). APE was not associated with a statistical significant increased risk on CRM involvement (OR ¼ 0.99; 95% CI ¼ 0.64e1.52) compared to LAR.
Characteristics
All patients N (%)
Total 1833 Sex Male 1148 Female 685 Age by subgroup 18e49 years 122 50e64 years 622 65þ years 1089 Number of comorbid conditions None 599 1 513 2þ 466 Unknown 255 Tumor distance to the anal verge Lower-rectum 558 Mid-rectum 616 Upper-rectum 520 Unknown 139 cT stage cT1 36 cT2 308 cT3 797 cT4 185 cTx 507 cN stage cN0 806 cN1 425 cN2 249 cNx 353 Neoadjuvant therapy 5 5 Gy RT 953 Chemoradiation 678 None 202 Operation technique LAR 969 APE 655 Hartman 150 Other 59 Surgical approach Open 1304 Laparoscopic 461 Conversion 68
CRM involved N (%)
CRM not involved N (%)
160 (9)
1673 (91)
101 (9) 59 (9)
1047 (91) 626 (91)
12 (10) 44 (7) 104 (10)
110 (90) 578 (93) 985 (90)
47 47 38 28
(8) (9) (8) (11)
552 466 428 227
(92) (91) (92) (89)
56 39 49 16
(10) (6) (9) (12)
502 577 471 123
(90) (94) (91) (89)
2 20 60 36 42
(6) (7) (8) (19) (8)
34 288 737 149 480
(94) (93) (93) (81) (92)
64 41 23 32
(8) (10) (9) (9)
742 384 226 321
(92) (90) (91) (91)
67 (7) 58 (9) 35 (17)
886 (93) 620 (91) 167 (83)
71 60 21 8
898 595 129 51
(7) (9) (14) (14)
130 (10) 28 (6) 2 (3)
(93) (91) (86) (86)
1174 (90) 433 (94) 66 (97)
CRM, circumferential resection margin; Gy, gray; RT, radiotherapy; LAR, low anterior resection; APE, abdomino perineal excision; cT, clinical T stage; cN clinical N stage.
Over the years an increase in CRM reporting was observed, from 77 to 90% accuracy of the pathology reports ( p < 0.001 [Fig. 1]). A large variation in CRM reporting was demonstrated, with missing reports ranging from 6% to 30% between the pathological departments ( p < 0.001 [Fig. 2]) and 5%e33% between the hospitals ( p < 0.001 [Fig. 3]). CRM involvement
Discussion The involvement of CRM is the most important factor for local recurrence and overall survival in the treatment of rectal cancer. Several risk factors are known for CRM involvement, such as higher pathological T and N stage. The present study reporting data from a large registry in Table 2 The use of neoadjuvant therapy plotted against clinical T-stage (cT). Clinical T-stage
5 5 Gy RT N (%)
CRT N (%)
No neoadjuvant therapy N (%)
Total N (%)
cT1 cT2 cT3 cT4 cTx Total
19 229 347 34 324 953
3 53 401 140 81 678
14 26 49 11 102 202
36 308 797 185 507 1833
(53) (74) (44) (18) (64) (52)
(8) (17) (50) (76) (16) (37)
(39) (8) (6) (6) (20) (11)
(2) (17) (44) (10) (28) (100)
GY, gray; RT, radiation therapy; CRT, chemoradiation therapy.
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
4
J. Homan et al. / EJSO xx (2015) 1e8
Figure 1. The reporting of CRM involvement by year of diagnosis. On the x-axis the year of diagnosis. Missing CRM reports are decreasing over the years, whereas CRM involvement decreased.
Figure 2. The reporting of CRM involvement by pathology laboratory. The different laboratories are listed on the x-axis. * Pathology department unknown or outside the ECR area.
the South of the Netherlands demonstrated that CRM reporting increased significantly to more than 90% in the recent years. However, significant differences remain in reporting CRM between pathology laboratories. Risk factors for involved CRM in multivariate analyses were cT4 stage, whereas neoadjuvant (chemo)radiation therapy lead to significantly less involved CRM. Furthermore, the type of operation (LAR or APE) did no longer significantly influence the risk on CRM involvement. Although there are no clinical consequences for not reporting CRM in the individual patient, it is of paramount importance for predicting an increased risk of local recurrence.1,2,4,12,17 CRM reporting also provides essential feedback to the surgeon and makes it possible to perform population-based analysis of the quality of rectal cancer surgery. In the present study CRM reporting increased
significantly, but in 2012, still a number (10%) of pathology reports do not contain this essential information. An important finding was the difference in CRM reporting between the six pathology laboratories. This was previously described by Bull et al.18 in Wales, where CRM reporting ranged from 8% in the worst to 78% in the best performing pathology departments in 1993. An explanation for this difference in reporting might be the lack of a standardized proforma for histopathology reporting of rectal cancer.19,20 Beattie et al.19 showed an absolute increase in reporting of CRM, from 79% to 95% after standardized proformas were introduced. Next to the introduction of the standardized proformas, the Dutch Surgical Colorectal Audit (DSCA) started with clinically auditing the outcomes of colorectal cancer surgery in the Netherlands. Due to these changes a further increase of CRM reporting is expected.15
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
J. Homan et al. / EJSO xx (2015) 1e8
5
Figure 3. The reporting of CRM involvement and missing cases per hospital. On the x-axis the different hospitals. * Hospital of surgery unknown or operated outside the ECR area. Table 3 Multivariable logistic regression analysis for circumferential resection margin involvement. Variables Sex Male Female Age 18e49 years old 50e64 years old >65 years old Clinical T-stage cT1/2 cT3 cT4 cTx Clinical N-stage cN0 cN1 cN2 cNx Neoadjuvant therapy None 5 5 Gy RT Chemoradiation Distance to anal verge Upper rectum Mid rectum Lower rectum Unknown Type of surgery LAR APE Hartman Other Surgical approach Open Laparoscopy Conversion
Odds ratio
CI 95% lowehigh
1.00 0.72
0.66e1.33
1.00 0.71 0.89
0.36e1.40 0.46e1.72
1.00 1.26 3.51 1.13
0.74e2.16 1.85e6.65 0.65e1.97
1.00 1.41 1.47 1.11
0.90e2.22 0.82e2.62 0.70e1.76
1.00 0.39 0.30
0.25e0.62 0.17e0.53
1.00 0.73 1.22 1.12
0.46e1.15 0.75e1.98 0.59e2.14
1.00 0.99 1.60 1.08
0.64e1.52 0.92e2.76 0.47e2.51
1.00 0.65 0.30
0.42e1.02 0.07e1.26
The variables that are bold in the text are statistical significant. LAR, low anterior resection; APE, abdomino perineal excision; Gy, gray; RT, radiotherapy.
CRM involvement has significantly decreased over the last decades, which is illustrated by a low overall CRM involvement of 10% in the present population-based study. In the last decade of the 20th century CRM involvement ranged from 12 to 30%12,17 and decreased in the beginning of the new millennium to 8e18% involvement.2,11,21,22 Most of these reports are from dedicated centers or randomized trials, which makes the excellent results of the present study of unselected patients in a registry remarkably good. This decrease might be explained by better preoperative screening of advanced disease and improvements in TME surgery.23 According to several guidelines since 2000 preoperative MRI is essential to provide vital information on the cT stage of the tumor, CRM involvement and suspicious lymph nodes.24 This might not only lead to a better surgical strategy to perform beyond TME surgery when necessary, but also leads to appropriate use of chemoradiation for downsizing locally advanced tumor.25 CRM involvement can also be considered as an indicator for quality of surgery of an individual surgeon or a hospital.3,12 In the present study univariate significant variations in CRM involvement and reporting was found between the 10 hospitals. Several other studies have analyzed the association between higher hospital operation volume and CRM involvement26 and in a previous Dutch study hospitals with low volume delivered a similar quality of surgery in terms of CRM involvement compared to the high volume hospitals. In the present study, however, cT4 tumors were excluded when comparing hospitals since these locally advanced rectal cancer (LARC) are generally referred to high volume, specialized hospitals. Our data demonstrated a non-significant trend towards an increased risk on CRM involvement in the low volume hospital. However, numbers of patients were small in this series and hampered case mix adjusted analyses.
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
6
J. Homan et al. / EJSO xx (2015) 1e8
Higher cT stage was associated with a higher rate of CRM involvement, which has been described previously for both clinical and pathological T staging.11,12,27,28 Also pathologically proven N2 stage was associated with high CRM involvement (data not shown). West et al.10 previously explained this by the fact that pN2 stage rectal cancer is related with more extensive primary tumors (T3 or T4 stage) and therefore with a higher incidence of CRM involvement. In the current study a higher rate of CRM involvement could not be established for cN2 stage, which is possibly due to difficult interpretation of preoperative MRI. A relatively high number of patients in the present study underwent some form of neoadjuvant radiotherapy, even in early staged rectal cancer patients. Reason for this was the Dutch National guidelines advising to treat all rectal cancer patients (besides cT1 tumors) with short term 5 5 Gy RT without waiting time to surgery. In the multivariable analysis both CRT and short term 5 5 Gy RT were associated with a decreased risk of CRM involvement. However, Nagtegaal et al.29 demonstrated that downstaging or downsizing does not occur in the 5 5 Gy regimen. A possible explanation of the low CRM involvement in patients treated with short course radiotherapy might be that patients undergoing emergency surgery, due to a large obstructing tumor, did not receive any neoadjuvant therapy, although data on this was not available. CRT on the other hand, is used for patients with locally advanced disease to enable downsizing and downstaging.30,31 It is an important finding that this population-based study reported excellent results in patients. In the Dutch TME study there was a significant difference in CRM involvement between LAR and APE, respectively 14% and 29%.2 This significant difference was also reported in several other studies.11,12,32 In the multivariate analysis of the present study the surgical technique was no longer a predictor for CRM involvement. This last finding is in line with recent publications demonstrating improved results in patients treated with APE.15,23,32 Factors that might explain this better outcome are improved preoperative imaging (i.e. MRI), introduction of new operation techniques (i.e. ELAPE) and more dedicated colorectal surgeons.33,34 A previous study by Tekkis et al. demonstrated that patients who underwent a Hartmann’s procedure have a higher rate of CRM involvement.35 The authors suggested this was because this operation is being performed more frequently in a palliative setting, dictated by more extensive tumors. In the present study a small group of patients underwent a Hartmann’s procedure and had a higher, but not statistically significant involved CRM. Since the introduction of laparoscopic surgery many studies were performed comparing laparoscopic surgery versus open surgery for rectal cancer.36e38 Laparoscopic surgery has advantages of earlier postoperative recovery, less blood loss, and lower rates of adhesion-related bowel obstruction. In addition, oncological outcome is comparable after laparoscopic and open surgery for rectal cancer. In the present study no significant difference on CRM
involvement was found between laparoscopic versus open surgery. Recently, Bonjer et al. also demonstrated this in a randomized trial comparing laparoscopic and open surgery with also similar long term results in locoregional recurrence and overall survival.39 Despite the introduction and development of the above mentioned factors the CRM was involved in 47% of the pathological T4 tumors (data not shown). An important reason for this high CRM involvement might be the fact that only 68% of the patients with a pathological T4 tumors were clinically staged as T4 tumors. This underestimation of clinical T stage can influence the type of neoadjuvant therapy that is given, but might also influence the surgical technique. Instead of beyond TME surgery, which should be performed in T4 tumors, patients are potentially operated with standard surgical techniques. Introduction of standard MRI after neoadjuvant therapy and more dedicated radiologists might be a solution to reduce the change on discrepancy between clinical and pathological T stage.40 In conclusion, the present population-based study demonstrated a high variation in reporting of CRM involvement between different hospitals and pathology departments, although significant improvements were made during the last years. CRM involvement is no longer dependent on the location of the tumor or the surgical procedure, but T stage and neo-adjuvant treatment are important factors predicting CRM involvement. Conflict of interest The authors declare that they have no conflict of interest. Funding This research did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector. Acknowledgments We would like to thank the participating hospitals of the Comprehensive Cancer Center South: Amphia Hospital, Breda; VieCuri Hospital, Venlo; Bernhoven Hospital, Veldhoven; St. Anna Hospital, Geldrop and Eindhoven; Catherina Hospital, Eindhoven; Elkerliek Hospital, Helmond; Maxima Medical Centre, Eindhoven and Veldhoven; St. Elisabeth Hospital, Tilburg; Jeroen Bosch Hospital, Den Bosch and Twee Steden Hospital, Tilburg. References 1. Nagtegaal ID, Quirke P. What is the role for the circumferential margin in the modern treatment of rectal cancer? J Clin Oncol 2008 Jan 10;26(2):303–12. PubMed PMID: 18182672. 2. Nagtegaal ID, Marijnen CA, Kranenbarg EK, et al. Circumferential margin involvement is still an important predictor of local recurrence
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
J. Homan et al. / EJSO xx (2015) 1e8
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
in rectal carcinoma: not one millimeter but two millimeters is the limit. Am J Surg Pathol 2002 Mar;26(3):350–7. PubMed PMID: 11859207. Quirke P, Durdey P, Dixon MF, Williams NS. Local recurrence of rectal adenocarcinoma due to inadequate surgical resection. Histopathological study of lateral tumour spread and surgical excision. Lancet 1986 Nov 1;2(8514):996–9. PubMed PMID: 2430152. den Dulk M, Krijnen P, Marijnen CA, et al. Improved overall survival for patients with rectal cancer since 1990: the effects of TME surgery and pre-operative radiotherapy. Eur J Cancer 2008 Aug;44(12):1710– 6. PubMed PMID: 18573654. Costa-Silva L, Brown G. Magnetic resonance imaging of rectal cancer. Magn Reson Imaging Clin N Am 2013 May;21(2):385–408. PubMed PMID: 23642559. Beets-Tan RG, Beets GL. Local staging of rectal cancer: a review of imaging. J Magn Reson Imaging 2011 May;33(5):1012–9. PubMed PMID: 21509856. Sebag-Montefiore D, Stephens RJ, Steele R, et al. Preoperative radiotherapy versus selective postoperative chemoradiotherapy in patients with rectal cancer (MRC CR07 and NCIC-CTG C016): a multicentre, randomised trial. Lancet 2009 Mar 7;373(9666):811–20. PubMed PMID: 19269519. Pubmed Central PMCID: 2668947. Kapiteijn E, Marijnen CA, Nagtegaal ID, et al. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med 2001 Aug 30;345(9):638–46. PubMed PMID: 11547717. Sauer R, Becker H, Hohenberger W, et al. Preoperative versus postoperative chemoradiotherapy for rectal cancer. N Engl J Med 2004 Oct 21;351(17):1731–40. PubMed PMID: 15496622. West NP, Anderin C, Smith KJ, Holm T, Quirke P. European Extralevator Abdominoperineal Excision Study G. Multicentre experience with extralevator abdominoperineal excision for low rectal cancer. Br J Surg 2010 Apr;97(4):588–99. PubMed PMID: 20186891. Baik SH, Kim NK, Lee YC, et al. Prognostic significance of circumferential resection margin following total mesorectal excision and adjuvant chemoradiotherapy in patients with rectal cancer. Ann Surg Oncol 2007 Feb;14(2):462–9. PubMed PMID: 17096053. Birbeck KF, Macklin CP, Tiffin NJ, et al. Rates of circumferential resection margin involvement vary between surgeons and predict outcomes in rectal cancer surgery. Ann Surg 2002 Apr;235(4): 449–57. PubMed PMID: 11923599. Pubmed Central PMCID: 1422458. Martijnse IS, Dudink RL, West NP, et al. Focus on extralevator perineal dissection in supine position for low rectal cancer has led to better quality of surgery and oncologic outcome. Ann Surg Oncol 2012 Mar; 19(3):786–93. PubMed PMID: 21861224. Stelzner S, Koehler C, Stelzer J, Sims A, Witzigmann H. Extended abdominoperineal excision vs. standard abdominoperineal excision in rectal cancerea systematic overview. Int J Colorectal Dis 2011 Oct; 26(10):1227–40. PubMed PMID: 21603901. van Leersum N, Martijnse I, den Dulk M, et al. Differences in circumferential resection margin involvement after abdominoperineal excision and low anterior resection no longer significant. Ann Surg 2014 Jun;259(6):1150–5. PubMed PMID: 24096756. Schouten LJ, Hoppener P, van den Brandt PA, Knottnerus JA, Jager JJ. Completeness of cancer registration in Limburg, The Netherlands. Int J Epidemiol 1993 Jun;22(3):369–76. PubMed PMID: 8359950. Adam IJ, Mohamdee MO, Martin IG, et al. Role of circumferential margin involvement in the local recurrence of rectal cancer. Lancet 1994 Sep 10;344(8924):707–11. PubMed PMID: 7915774. Bull AD, Biffin AH, Mella J, et al. Colorectal cancer pathology reporting: a regional audit. J Clin Pathol 1997 Feb;50(2):138–42. PubMed PMID: 9155695. Pubmed Central PMCID: 499739. Beattie GC, McAdam TK, Elliott S, Sloan JM, Irwin ST. Improvement in quality of colorectal cancer pathology reporting with a standardized proformaea comparative study. Colorectal Dis 2003 Nov;5(6):558– 62. PubMed PMID: 14617240.
7
20. Greening SEBL, Newcombe R. A randomized controlled trial of the value of guidelines and computerized proformas in pathology reporting of cancer specimans: the Wales CROPS project. J Pathol 2000; 190(31A). 21. Khani MH, Smedh K, Kraaz W. Is the circumferential resection margin a predictor of local recurrence after preoperative radiotherapy and optimal surgery for rectal carcinoma? Colorectal Dis 2007 Oct; 9(8):706–12. PubMed PMID: 17535279. 22. Wibe A, Rendedal PR, Svensson E, et al. Prognostic significance of the circumferential resection margin following total mesorectal excision for rectal cancer. Br J Surg 2002 Mar;89(3):327–34. PubMed PMID: 11872058. 23. Bokkerink GM, Buijs EF, de Ruijter W, et al. Improved quality of care for patients undergoing an abdominoperineal excision for rectal cancer. Eur J Surg Oncol 2015 Feb;41(2):201–7. PubMed PMID: 25572974. 24. Salerno GV, Daniels IR, Moran BJ, Heald RJ, Thomas K, Brown G. Magnetic resonance imaging prediction of an involved surgical resection margin in low rectal cancer. Dis Colon Rectum 2009 Apr;52(4): 632–9. PubMed PMID: 19404067. 25. Beyond TMEC. Consensus statement on the multidisciplinary management of patients with recurrent and primary rectal cancer beyond total mesorectal excision planes. Br J Surg 2013 Jul;100(8):E1–E33. PubMed PMID: 23901427. 26. van Erning FN, van Steenbergen LN, van den Broek WT, Rutten HJ, Lemmens VE. No difference between lowest and highest volume hospitals in outcome after colorectal cancer surgery in the southern Netherlands. Eur J Surg Oncol 2013 Nov;39(11):1199–206. PubMed PMID: 24044806. 27. Wu ZY, Wan J, Li JH, et al. Prognostic value of lateral lymph node metastasis for advanced low rectal cancer. World J Gastroenterol 2007 Dec 7;13(45):6048–52. PubMed PMID: 18023098. 28. Gosens MJ, Klaassen RA, Tan-Go I, et al. Circumferential margin involvement is the crucial prognostic factor after multimodality treatment in patients with locally advanced rectal carcinoma. Clin Cancer Res 2007 Nov 15;13(22 Pt 1):6617–23. PubMed PMID: 18006762. 29. Marijnen CA, Nagtegaal ID, Klein Kranenbarg E, et al. No downstaging after short-term preoperative radiotherapy in rectal cancer patients. J Clin Oncol 2001 Apr 1;19(7):1976–84. PubMed PMID: 11283130. 30. Wolthuis AM, Penninckx F, Haustermans K, et al. Impact of interval between neoadjuvant chemoradiotherapy and TME for locally advanced rectal cancer on pathologic response and oncologic outcome. Ann Surg Oncol 2012 Sep;19(9):2833–41. PubMed PMID: 22451236. 31. Evans J, Tait D, Swift I, et al. Timing of surgery following preoperative therapy in rectal cancer: the need for a prospective randomized trial? Dis Colon Rectum 2011 Oct;54(10):1251–9. PubMed PMID: 21904139. 32. Shihab OC, Brown G, Daniels IR, Heald RJ, Quirke P, Moran BJ. Patients with low rectal cancer treated by abdominoperineal excision have worse tumors and higher involved margin rates compared with patients treated by anterior resection. Dis Colon Rectum 2010 Jan; 53(1):53–6. PubMed PMID: 20010351. 33. Holm T, Ljung A, Haggmark T, Jurell G, Lagergren J. Extended abdominoperineal resection with gluteus maximus flap reconstruction of the pelvic floor for rectal cancer. Br J Surg 2007 Feb;94(2):232– 8. PubMed PMID: 17143848. 34. Martling A, Cedermark B, Johansson H, Rutqvist LE, Holm T. The surgeon as a prognostic factor after the introduction of total mesorectal excision in the treatment of rectal cancer. Br J Surg 2002 Aug;89(8): 1008–13. PubMed PMID: 12153626. 35. Tekkis PP, Heriot AG, Smith J, et al. Comparison of circumferential margin involvement between restorative and nonrestorative resections for rectal cancer. Colorectal Dis 2005 Jul;7(4):369–74. PubMed PMID: 15932561. 36. Green BL, Marshall HC, Collinson F, et al. Long-term follow-up of the Medical Research Council CLASICC trial of conventional versus
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
8
J. Homan et al. / EJSO xx (2015) 1e8
laparoscopically assisted resection in colorectal cancer. Br J Surg 2013 Jan;100(1):75–82. PubMed PMID: 23132548. 37. Ng SS, Leung KL, Lee JF, et al. Laparoscopic-assisted versus open abdominoperineal resection for low rectal cancer: a prospective randomized trial. Ann Surg Oncol 2008 Sep;15(9):2418–25. PubMed PMID: 18392659. 38. van der Pas MH, Haglind E, Cuesta MA, et al. Laparoscopic versus open surgery for rectal cancer (COLOR II): short-term outcomes of a randomised, phase 3 trial. Lancet Oncol 2013 Mar;14(3):210–8. PubMed PMID: 23395398.
39. Bonjer HJ, Deijen CL, Abis GA, et al. A randomized trial of laparoscopic versus open surgery for rectal cancer. N Engl J Med 2015 Apr 2;372(14):1324–32. PubMed PMID: 25830422. 40. Kulkarni T, Gollins S, Maw A, Hobson P, Byrne R, Widdowson D. Magnetic resonance imaging in rectal cancer downstaged using neoadjuvant chemoradiation: accuracy of prediction of tumour stage and circumferential resection margin status. Colorectal Dis 2008 Jun;10(5):479–89. PubMed PMID: 18318754.
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
ScienceDirect EJSO xx (2015) 1e8
www.ejso.com
Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands okkerink a, M.J. Aarts b, V.E. Lemmens b, J. Homan a,*, G.M. B€ G. van Lijnschoten c, H.J. Rutten d,e, J.H. Wijsman f, I.D. Nagtegaal g, J.H.W. de Wilt a a
Department of Surgery, Radboud University Medical Center Nijmegen, Nijmegen, The Netherlands b Netherlands Comprehensive Cancer Organization, Eindhoven, The Netherlands c Department of Pathology, PAMM, Eindhoven, The Netherlands d Department of Surgery, Catharina Hospital, Eindhoven, The Netherlands e Department of Surgery, Maastricht University Medical Center, Maastricht, The Netherlands f Department of Surgery, Amphia Hospital, Breda, The Netherlands g Department of Pathology, Radboud University Medical Center, Nijmegen, The Netherlands Accepted 9 July 2015 Available online - - -
Abstract Background: Since the introduction of total mesorectal surgery the outcome of rectal cancer patients has improved significantly. Involvement of the circumferential resection margin (CRM) is an important predictor of increased local recurrence, distant metastases and decreased overall survival. Abdomino perineal excision (APE) is associated with increased risk of CRM involvement. Aim of this study was to analyze reporting of CRM and to identify predictive factors for CRM involvement. Methods: A population-based dataset was used selecting 2153 patients diagnosed between 2008 and 2013 with primary rectal cancer undergoing surgery. Variation in CRM reporting was assessed and predictive factors for CRM involvement were calculated and used in multivariate analyses. Results: Large variation in CRM reporting was found between pathology departments, with missing cases varying from 6% to 30%. CRM reporting increased from 77% in 2008 to 90% in 2012 ( p < 0.001). CRM involvement significantly decreased from 12% to 6% over the years ( p < 0.001). In multivariate analysis type of operation, low anterior resection or APE, did not influence the risk of CRM involvement. Clinical T4stage [odds ratio (OR) ¼ 3.51; 95% confidence interval (CI) ¼ 1.85e6.65) was associated with increased risk of CRM involvement, whereas neoadjuvant treatment (5 5 gray radiotherapy [OR 0.39; CI 0.25e0.62] or chemoradiation therapy [OR 0.30; CI 0.17e0.53]) were associated with significant decreased risk of CRM involvement. Conclusion: Although significant improvements are made during the last years there still is variation in reporting of CRM involvement in the Southern Netherlands. In multivariate analysis APE was no longer associated with increased risk of CRM involvement. Ó 2015 Elsevier Ltd. All rights reserved.
Keywords: Rectal cancer; Circumferential resection margin (CRM); Involvement; Abdomino perineal excision (APE); Low anterior resection (LAR)
Introduction The circumferential resection margin (CRM) is an important prognostic factor in rectal cancer care and a * Corresponding author. Department of Surgery, Radboud University Medical Center Nijmegen, Geert Grooteplein-Zuid 10, 6525 GA, Nijmegen, The Netherlands. Tel.: þ31 0 24 361 1111. E-mail address: [email protected] (J. Homan).
predictor of local recurrence, distant metastases and overall survival.1,2 The CRM was first described in 1986 by Quirke3 and is part of a standardized histopathological protocol. Transverse sectioning of the excision specimen is one of the key procedures in this protocol, which is now recommended in almost every national rectal cancer guideline in Europe.
http://dx.doi.org/10.1016/j.ejso.2015.07.015 0748-7983/Ó 2015 Elsevier Ltd. All rights reserved. Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
2
J. Homan et al. / EJSO xx (2015) 1e8
Since the introduction of the CRM, the prognosis of patients with rectal cancer has dramatically improved. The introduction of total mesorectal excision (TME) and extralevator techniques for lower rectal tumors have led to a decrease in CRM involvement, and consequently to less local recurrence and improved metastasis free and overall survival.4 Besides improved surgical techniques, the introduction of standard magnetic resonance imaging (MRI) has enabled clinicians to properly select patients with locally advanced disease and treat them with preoperatively with (chemo)radiation therapy.5,6 Also, short course radiotherapy has proved to decrease local recurrence rates,7,8 similar to long course chemoradiation therapy for locally advanced tumors.9 Besides prognostic information, the CRM provides the surgeon of essential feedback on the quality of surgery.3 Determination of predictive factors for CRM involvement is essential to provide the best patient care. According to the literature more extensive tumors (i.e. T4 and N2 stage tumors) are related to higher CRM involvement rates.10e12 The results of the Dutch TME trial illustrated a significant difference in CRM involvement between low anterior resection (LAR) and abdomino perineal excision (APE), respectively 14% and 29% ( p < 0.001).2 On the other hand, the introduction of new surgical techniques, such as extralevator APE (ELAPE), resulted in less CRM involvement and better oncological outcome in patients with distal rectal cancer.13,14 Population-based data on CRM reporting and involvement are rare in the available literature.15 The aim of the current study was to evaluate variation in CRM reporting within different pathology departments and CRM involvement in different hospitals. Moreover, prognostic factors for CRM involvement were identified using the population-based database of the Eindhoven Cancer Registry (ECR). Patients and methods All patients diagnosed with primary rectal cancer (stage IeIII) between January 2008 and January 2013 were selected using population-based data from the Eindhoven Cancer Registry (ECR) which is part of the Netherlands Cancer Registry and maintained by the Netherlands Comprehensive Cancer Organization (IKNL). The ECR collects data on all newly diagnosed patients with rectal cancer in the Southern Netherlands. The ECR covers an area with ten community hospitals and six pathology departments. Due to intensive training of the registrars and computerized consistency checks, the quality of the retrospective data is high.16 Consent was obtained from all patients in accordance with the local and international legislation (declaration of Helsinki). Patients who were diagnosed in another region, but operated in one of the ECR hospitals were not included, as well as patients who did not undergo an operation. Other exclusion criteria were stage IV rectal cancer and patients who underwent a local excision of their rectal tumor.
The outcome of interest was CRM reporting and CRM involvement. Rectal cancer was defined as a carcinoma of the rectum within 15 cm of the anal verge. In the present study a negative or free CRM was defined as more than 1 mm, or when “free” was reported in the pathology report. A positive resection margin was defined as a margin of 1 mm or less, or when “positive” was reported in the pathology report. The CRM reporting was assessed by the year of diagnosis as well as for pathological laboratories of the ECR area. Other variables that were collected for analyses were: comorbidity according to the Charlsons comorbidity index, tumor characteristics, neoadjuvant therapy, type of surgery, and localization of the tumor (topography). Neoadjuvant therapy was divided in four groups; no neoadjuvant therapy, short term 5 5 gray (Gy) radiotherapy (RT), chemoradiation therapy (CRT), and only chemotherapy. The type of surgery was divided into four groups: the LAR, APE, Hartmann operations and other surgery (which included exenterations and proctocolectomies). The localization of the tumor was categorized into three groups (based on MRI if available): the lower-rectum (distance from anus 0e49 mm), the midrectum (distance from anus 50e99 mm) and the upperrectum (distance from anus 100 mm). For describing clinical tumor characteristics the following abbreviations are used; clinical tumor stage (cT), clinical node stage (cN). In the analysis to determine predictive factors for CRM involvement all patients with missing pathology reports were excluded. In the multivariate logistic regression analyses for CRM involvement adjustments were made for sex, age category (18e49, 50e64, or age more than 65 years old), cT, cN, the use of neoadjuvant thereapy (no therapy, 5 5 Gy RT, or CRT), distance of the tumor to the anal verge (lower, mid, and upper rectum), type of surgery (LAR, APE, Hartmann, or any other type of surgery), and surgical approach (open, laparoscopy, or conversion to open surgery). The derived odds ratios (ORs) and the 95% confidence intervals (CIs) are presented. Furthermore, it was evaluated whether the number of operations on rectal cancer per hospital per year influenced CRM involvement. To do so, hospitals were divided into three groups: hospitals in which 20 operations, 21 to 40 operations or >40 operations per year were performed. Because of centralization of cT4-stage tumors to specialized centers they were excluded from this analysis to avoid negative selection bias. All statistical analyses were performed using SPSS version 21.0 for Windows (Inc., Chicago, Illinois, USA). P-values were two-sided and values < 0.05 were considered significant. Results Patients and characteristics A total of 3348 patients were diagnosed with rectal cancer during the study period. In total 71% of the patients
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
J. Homan et al. / EJSO xx (2015) 1e8
3
were surgically treated with (beyond) TME surgery, and after exclusion of 224 stage IV patients a total of 2153 patients were included. In 1833 (85%) patients the CRM was reported in the pathological report. Mean age at time of surgery was 66 11 years and 63% of the patients were male. Table 1 gives an overview of different patient and tumor characteristics and CRM involvement rates. In total 89% of the patients received neoadjuvant therapy; 52% received 5 5 Gy RT and 37% CRT. In early tumors (cT1 and cT2) CRT was only used in 16% of the patients, whereas the use of CRT was 76% for cT4 tumors (Table 2).
CRM reporting
Table 1 The baseline characteristics of patients with rectal cancer registered in the ECR area. All patients with an unreported CRM in the pathology report were excluded from this analysis.
CRM involvement rate decreased significantly over the years, from 12% in 2008 to 6% in 2012 ( p < 0.001). In total 87% of the operations were performed within the ECR area, in the other 13% the hospital was unknown or not affiliated to the ECR area. The mean number of operations for rectal cancer ranged between the hospitals from 5 to 68 operations per year. There was one hospital in which less than 20 operations for rectal cancer per year were performed. After exclusion of the cT4 tumors there was a non-significant trend towards an increased CRM involvement of 13% in that hospital compared to 6% CRM involvement in the hospital in which more than 40 operations per year were performed ( p ¼ 0.069). The multivariate logistic regression analysis for CRM involvement is presented in Table 3. A significantly decreased risk of CRM involvement was observed for patients who received neoadjuvant CRT (OR ¼ 0.30; 95% CI ¼ 0.17e0.53) or 5 5 Gy RT (OR ¼ 0.39; 95% CI ¼ 0.25e0.62) compared to patients who did not receive neoadjuvant therapy. Clinical T4 stage was associated with an increased risk on CRM involvement (OR ¼ 3.51; 95% CI ¼ 1.85e6.65). APE was not associated with a statistical significant increased risk on CRM involvement (OR ¼ 0.99; 95% CI ¼ 0.64e1.52) compared to LAR.
Characteristics
All patients N (%)
Total 1833 Sex Male 1148 Female 685 Age by subgroup 18e49 years 122 50e64 years 622 65þ years 1089 Number of comorbid conditions None 599 1 513 2þ 466 Unknown 255 Tumor distance to the anal verge Lower-rectum 558 Mid-rectum 616 Upper-rectum 520 Unknown 139 cT stage cT1 36 cT2 308 cT3 797 cT4 185 cTx 507 cN stage cN0 806 cN1 425 cN2 249 cNx 353 Neoadjuvant therapy 5 5 Gy RT 953 Chemoradiation 678 None 202 Operation technique LAR 969 APE 655 Hartman 150 Other 59 Surgical approach Open 1304 Laparoscopic 461 Conversion 68
CRM involved N (%)
CRM not involved N (%)
160 (9)
1673 (91)
101 (9) 59 (9)
1047 (91) 626 (91)
12 (10) 44 (7) 104 (10)
110 (90) 578 (93) 985 (90)
47 47 38 28
(8) (9) (8) (11)
552 466 428 227
(92) (91) (92) (89)
56 39 49 16
(10) (6) (9) (12)
502 577 471 123
(90) (94) (91) (89)
2 20 60 36 42
(6) (7) (8) (19) (8)
34 288 737 149 480
(94) (93) (93) (81) (92)
64 41 23 32
(8) (10) (9) (9)
742 384 226 321
(92) (90) (91) (91)
67 (7) 58 (9) 35 (17)
886 (93) 620 (91) 167 (83)
71 60 21 8
898 595 129 51
(7) (9) (14) (14)
130 (10) 28 (6) 2 (3)
(93) (91) (86) (86)
1174 (90) 433 (94) 66 (97)
CRM, circumferential resection margin; Gy, gray; RT, radiotherapy; LAR, low anterior resection; APE, abdomino perineal excision; cT, clinical T stage; cN clinical N stage.
Over the years an increase in CRM reporting was observed, from 77 to 90% accuracy of the pathology reports ( p < 0.001 [Fig. 1]). A large variation in CRM reporting was demonstrated, with missing reports ranging from 6% to 30% between the pathological departments ( p < 0.001 [Fig. 2]) and 5%e33% between the hospitals ( p < 0.001 [Fig. 3]). CRM involvement
Discussion The involvement of CRM is the most important factor for local recurrence and overall survival in the treatment of rectal cancer. Several risk factors are known for CRM involvement, such as higher pathological T and N stage. The present study reporting data from a large registry in Table 2 The use of neoadjuvant therapy plotted against clinical T-stage (cT). Clinical T-stage
5 5 Gy RT N (%)
CRT N (%)
No neoadjuvant therapy N (%)
Total N (%)
cT1 cT2 cT3 cT4 cTx Total
19 229 347 34 324 953
3 53 401 140 81 678
14 26 49 11 102 202
36 308 797 185 507 1833
(53) (74) (44) (18) (64) (52)
(8) (17) (50) (76) (16) (37)
(39) (8) (6) (6) (20) (11)
(2) (17) (44) (10) (28) (100)
GY, gray; RT, radiation therapy; CRT, chemoradiation therapy.
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
4
J. Homan et al. / EJSO xx (2015) 1e8
Figure 1. The reporting of CRM involvement by year of diagnosis. On the x-axis the year of diagnosis. Missing CRM reports are decreasing over the years, whereas CRM involvement decreased.
Figure 2. The reporting of CRM involvement by pathology laboratory. The different laboratories are listed on the x-axis. * Pathology department unknown or outside the ECR area.
the South of the Netherlands demonstrated that CRM reporting increased significantly to more than 90% in the recent years. However, significant differences remain in reporting CRM between pathology laboratories. Risk factors for involved CRM in multivariate analyses were cT4 stage, whereas neoadjuvant (chemo)radiation therapy lead to significantly less involved CRM. Furthermore, the type of operation (LAR or APE) did no longer significantly influence the risk on CRM involvement. Although there are no clinical consequences for not reporting CRM in the individual patient, it is of paramount importance for predicting an increased risk of local recurrence.1,2,4,12,17 CRM reporting also provides essential feedback to the surgeon and makes it possible to perform population-based analysis of the quality of rectal cancer surgery. In the present study CRM reporting increased
significantly, but in 2012, still a number (10%) of pathology reports do not contain this essential information. An important finding was the difference in CRM reporting between the six pathology laboratories. This was previously described by Bull et al.18 in Wales, where CRM reporting ranged from 8% in the worst to 78% in the best performing pathology departments in 1993. An explanation for this difference in reporting might be the lack of a standardized proforma for histopathology reporting of rectal cancer.19,20 Beattie et al.19 showed an absolute increase in reporting of CRM, from 79% to 95% after standardized proformas were introduced. Next to the introduction of the standardized proformas, the Dutch Surgical Colorectal Audit (DSCA) started with clinically auditing the outcomes of colorectal cancer surgery in the Netherlands. Due to these changes a further increase of CRM reporting is expected.15
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
J. Homan et al. / EJSO xx (2015) 1e8
5
Figure 3. The reporting of CRM involvement and missing cases per hospital. On the x-axis the different hospitals. * Hospital of surgery unknown or operated outside the ECR area. Table 3 Multivariable logistic regression analysis for circumferential resection margin involvement. Variables Sex Male Female Age 18e49 years old 50e64 years old >65 years old Clinical T-stage cT1/2 cT3 cT4 cTx Clinical N-stage cN0 cN1 cN2 cNx Neoadjuvant therapy None 5 5 Gy RT Chemoradiation Distance to anal verge Upper rectum Mid rectum Lower rectum Unknown Type of surgery LAR APE Hartman Other Surgical approach Open Laparoscopy Conversion
Odds ratio
CI 95% lowehigh
1.00 0.72
0.66e1.33
1.00 0.71 0.89
0.36e1.40 0.46e1.72
1.00 1.26 3.51 1.13
0.74e2.16 1.85e6.65 0.65e1.97
1.00 1.41 1.47 1.11
0.90e2.22 0.82e2.62 0.70e1.76
1.00 0.39 0.30
0.25e0.62 0.17e0.53
1.00 0.73 1.22 1.12
0.46e1.15 0.75e1.98 0.59e2.14
1.00 0.99 1.60 1.08
0.64e1.52 0.92e2.76 0.47e2.51
1.00 0.65 0.30
0.42e1.02 0.07e1.26
The variables that are bold in the text are statistical significant. LAR, low anterior resection; APE, abdomino perineal excision; Gy, gray; RT, radiotherapy.
CRM involvement has significantly decreased over the last decades, which is illustrated by a low overall CRM involvement of 10% in the present population-based study. In the last decade of the 20th century CRM involvement ranged from 12 to 30%12,17 and decreased in the beginning of the new millennium to 8e18% involvement.2,11,21,22 Most of these reports are from dedicated centers or randomized trials, which makes the excellent results of the present study of unselected patients in a registry remarkably good. This decrease might be explained by better preoperative screening of advanced disease and improvements in TME surgery.23 According to several guidelines since 2000 preoperative MRI is essential to provide vital information on the cT stage of the tumor, CRM involvement and suspicious lymph nodes.24 This might not only lead to a better surgical strategy to perform beyond TME surgery when necessary, but also leads to appropriate use of chemoradiation for downsizing locally advanced tumor.25 CRM involvement can also be considered as an indicator for quality of surgery of an individual surgeon or a hospital.3,12 In the present study univariate significant variations in CRM involvement and reporting was found between the 10 hospitals. Several other studies have analyzed the association between higher hospital operation volume and CRM involvement26 and in a previous Dutch study hospitals with low volume delivered a similar quality of surgery in terms of CRM involvement compared to the high volume hospitals. In the present study, however, cT4 tumors were excluded when comparing hospitals since these locally advanced rectal cancer (LARC) are generally referred to high volume, specialized hospitals. Our data demonstrated a non-significant trend towards an increased risk on CRM involvement in the low volume hospital. However, numbers of patients were small in this series and hampered case mix adjusted analyses.
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
6
J. Homan et al. / EJSO xx (2015) 1e8
Higher cT stage was associated with a higher rate of CRM involvement, which has been described previously for both clinical and pathological T staging.11,12,27,28 Also pathologically proven N2 stage was associated with high CRM involvement (data not shown). West et al.10 previously explained this by the fact that pN2 stage rectal cancer is related with more extensive primary tumors (T3 or T4 stage) and therefore with a higher incidence of CRM involvement. In the current study a higher rate of CRM involvement could not be established for cN2 stage, which is possibly due to difficult interpretation of preoperative MRI. A relatively high number of patients in the present study underwent some form of neoadjuvant radiotherapy, even in early staged rectal cancer patients. Reason for this was the Dutch National guidelines advising to treat all rectal cancer patients (besides cT1 tumors) with short term 5 5 Gy RT without waiting time to surgery. In the multivariable analysis both CRT and short term 5 5 Gy RT were associated with a decreased risk of CRM involvement. However, Nagtegaal et al.29 demonstrated that downstaging or downsizing does not occur in the 5 5 Gy regimen. A possible explanation of the low CRM involvement in patients treated with short course radiotherapy might be that patients undergoing emergency surgery, due to a large obstructing tumor, did not receive any neoadjuvant therapy, although data on this was not available. CRT on the other hand, is used for patients with locally advanced disease to enable downsizing and downstaging.30,31 It is an important finding that this population-based study reported excellent results in patients. In the Dutch TME study there was a significant difference in CRM involvement between LAR and APE, respectively 14% and 29%.2 This significant difference was also reported in several other studies.11,12,32 In the multivariate analysis of the present study the surgical technique was no longer a predictor for CRM involvement. This last finding is in line with recent publications demonstrating improved results in patients treated with APE.15,23,32 Factors that might explain this better outcome are improved preoperative imaging (i.e. MRI), introduction of new operation techniques (i.e. ELAPE) and more dedicated colorectal surgeons.33,34 A previous study by Tekkis et al. demonstrated that patients who underwent a Hartmann’s procedure have a higher rate of CRM involvement.35 The authors suggested this was because this operation is being performed more frequently in a palliative setting, dictated by more extensive tumors. In the present study a small group of patients underwent a Hartmann’s procedure and had a higher, but not statistically significant involved CRM. Since the introduction of laparoscopic surgery many studies were performed comparing laparoscopic surgery versus open surgery for rectal cancer.36e38 Laparoscopic surgery has advantages of earlier postoperative recovery, less blood loss, and lower rates of adhesion-related bowel obstruction. In addition, oncological outcome is comparable after laparoscopic and open surgery for rectal cancer. In the present study no significant difference on CRM
involvement was found between laparoscopic versus open surgery. Recently, Bonjer et al. also demonstrated this in a randomized trial comparing laparoscopic and open surgery with also similar long term results in locoregional recurrence and overall survival.39 Despite the introduction and development of the above mentioned factors the CRM was involved in 47% of the pathological T4 tumors (data not shown). An important reason for this high CRM involvement might be the fact that only 68% of the patients with a pathological T4 tumors were clinically staged as T4 tumors. This underestimation of clinical T stage can influence the type of neoadjuvant therapy that is given, but might also influence the surgical technique. Instead of beyond TME surgery, which should be performed in T4 tumors, patients are potentially operated with standard surgical techniques. Introduction of standard MRI after neoadjuvant therapy and more dedicated radiologists might be a solution to reduce the change on discrepancy between clinical and pathological T stage.40 In conclusion, the present population-based study demonstrated a high variation in reporting of CRM involvement between different hospitals and pathology departments, although significant improvements were made during the last years. CRM involvement is no longer dependent on the location of the tumor or the surgical procedure, but T stage and neo-adjuvant treatment are important factors predicting CRM involvement. Conflict of interest The authors declare that they have no conflict of interest. Funding This research did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector. Acknowledgments We would like to thank the participating hospitals of the Comprehensive Cancer Center South: Amphia Hospital, Breda; VieCuri Hospital, Venlo; Bernhoven Hospital, Veldhoven; St. Anna Hospital, Geldrop and Eindhoven; Catherina Hospital, Eindhoven; Elkerliek Hospital, Helmond; Maxima Medical Centre, Eindhoven and Veldhoven; St. Elisabeth Hospital, Tilburg; Jeroen Bosch Hospital, Den Bosch and Twee Steden Hospital, Tilburg. References 1. Nagtegaal ID, Quirke P. What is the role for the circumferential margin in the modern treatment of rectal cancer? J Clin Oncol 2008 Jan 10;26(2):303–12. PubMed PMID: 18182672. 2. Nagtegaal ID, Marijnen CA, Kranenbarg EK, et al. Circumferential margin involvement is still an important predictor of local recurrence
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
J. Homan et al. / EJSO xx (2015) 1e8
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
in rectal carcinoma: not one millimeter but two millimeters is the limit. Am J Surg Pathol 2002 Mar;26(3):350–7. PubMed PMID: 11859207. Quirke P, Durdey P, Dixon MF, Williams NS. Local recurrence of rectal adenocarcinoma due to inadequate surgical resection. Histopathological study of lateral tumour spread and surgical excision. Lancet 1986 Nov 1;2(8514):996–9. PubMed PMID: 2430152. den Dulk M, Krijnen P, Marijnen CA, et al. Improved overall survival for patients with rectal cancer since 1990: the effects of TME surgery and pre-operative radiotherapy. Eur J Cancer 2008 Aug;44(12):1710– 6. PubMed PMID: 18573654. Costa-Silva L, Brown G. Magnetic resonance imaging of rectal cancer. Magn Reson Imaging Clin N Am 2013 May;21(2):385–408. PubMed PMID: 23642559. Beets-Tan RG, Beets GL. Local staging of rectal cancer: a review of imaging. J Magn Reson Imaging 2011 May;33(5):1012–9. PubMed PMID: 21509856. Sebag-Montefiore D, Stephens RJ, Steele R, et al. Preoperative radiotherapy versus selective postoperative chemoradiotherapy in patients with rectal cancer (MRC CR07 and NCIC-CTG C016): a multicentre, randomised trial. Lancet 2009 Mar 7;373(9666):811–20. PubMed PMID: 19269519. Pubmed Central PMCID: 2668947. Kapiteijn E, Marijnen CA, Nagtegaal ID, et al. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med 2001 Aug 30;345(9):638–46. PubMed PMID: 11547717. Sauer R, Becker H, Hohenberger W, et al. Preoperative versus postoperative chemoradiotherapy for rectal cancer. N Engl J Med 2004 Oct 21;351(17):1731–40. PubMed PMID: 15496622. West NP, Anderin C, Smith KJ, Holm T, Quirke P. European Extralevator Abdominoperineal Excision Study G. Multicentre experience with extralevator abdominoperineal excision for low rectal cancer. Br J Surg 2010 Apr;97(4):588–99. PubMed PMID: 20186891. Baik SH, Kim NK, Lee YC, et al. Prognostic significance of circumferential resection margin following total mesorectal excision and adjuvant chemoradiotherapy in patients with rectal cancer. Ann Surg Oncol 2007 Feb;14(2):462–9. PubMed PMID: 17096053. Birbeck KF, Macklin CP, Tiffin NJ, et al. Rates of circumferential resection margin involvement vary between surgeons and predict outcomes in rectal cancer surgery. Ann Surg 2002 Apr;235(4): 449–57. PubMed PMID: 11923599. Pubmed Central PMCID: 1422458. Martijnse IS, Dudink RL, West NP, et al. Focus on extralevator perineal dissection in supine position for low rectal cancer has led to better quality of surgery and oncologic outcome. Ann Surg Oncol 2012 Mar; 19(3):786–93. PubMed PMID: 21861224. Stelzner S, Koehler C, Stelzer J, Sims A, Witzigmann H. Extended abdominoperineal excision vs. standard abdominoperineal excision in rectal cancerea systematic overview. Int J Colorectal Dis 2011 Oct; 26(10):1227–40. PubMed PMID: 21603901. van Leersum N, Martijnse I, den Dulk M, et al. Differences in circumferential resection margin involvement after abdominoperineal excision and low anterior resection no longer significant. Ann Surg 2014 Jun;259(6):1150–5. PubMed PMID: 24096756. Schouten LJ, Hoppener P, van den Brandt PA, Knottnerus JA, Jager JJ. Completeness of cancer registration in Limburg, The Netherlands. Int J Epidemiol 1993 Jun;22(3):369–76. PubMed PMID: 8359950. Adam IJ, Mohamdee MO, Martin IG, et al. Role of circumferential margin involvement in the local recurrence of rectal cancer. Lancet 1994 Sep 10;344(8924):707–11. PubMed PMID: 7915774. Bull AD, Biffin AH, Mella J, et al. Colorectal cancer pathology reporting: a regional audit. J Clin Pathol 1997 Feb;50(2):138–42. PubMed PMID: 9155695. Pubmed Central PMCID: 499739. Beattie GC, McAdam TK, Elliott S, Sloan JM, Irwin ST. Improvement in quality of colorectal cancer pathology reporting with a standardized proformaea comparative study. Colorectal Dis 2003 Nov;5(6):558– 62. PubMed PMID: 14617240.
7
20. Greening SEBL, Newcombe R. A randomized controlled trial of the value of guidelines and computerized proformas in pathology reporting of cancer specimans: the Wales CROPS project. J Pathol 2000; 190(31A). 21. Khani MH, Smedh K, Kraaz W. Is the circumferential resection margin a predictor of local recurrence after preoperative radiotherapy and optimal surgery for rectal carcinoma? Colorectal Dis 2007 Oct; 9(8):706–12. PubMed PMID: 17535279. 22. Wibe A, Rendedal PR, Svensson E, et al. Prognostic significance of the circumferential resection margin following total mesorectal excision for rectal cancer. Br J Surg 2002 Mar;89(3):327–34. PubMed PMID: 11872058. 23. Bokkerink GM, Buijs EF, de Ruijter W, et al. Improved quality of care for patients undergoing an abdominoperineal excision for rectal cancer. Eur J Surg Oncol 2015 Feb;41(2):201–7. PubMed PMID: 25572974. 24. Salerno GV, Daniels IR, Moran BJ, Heald RJ, Thomas K, Brown G. Magnetic resonance imaging prediction of an involved surgical resection margin in low rectal cancer. Dis Colon Rectum 2009 Apr;52(4): 632–9. PubMed PMID: 19404067. 25. Beyond TMEC. Consensus statement on the multidisciplinary management of patients with recurrent and primary rectal cancer beyond total mesorectal excision planes. Br J Surg 2013 Jul;100(8):E1–E33. PubMed PMID: 23901427. 26. van Erning FN, van Steenbergen LN, van den Broek WT, Rutten HJ, Lemmens VE. No difference between lowest and highest volume hospitals in outcome after colorectal cancer surgery in the southern Netherlands. Eur J Surg Oncol 2013 Nov;39(11):1199–206. PubMed PMID: 24044806. 27. Wu ZY, Wan J, Li JH, et al. Prognostic value of lateral lymph node metastasis for advanced low rectal cancer. World J Gastroenterol 2007 Dec 7;13(45):6048–52. PubMed PMID: 18023098. 28. Gosens MJ, Klaassen RA, Tan-Go I, et al. Circumferential margin involvement is the crucial prognostic factor after multimodality treatment in patients with locally advanced rectal carcinoma. Clin Cancer Res 2007 Nov 15;13(22 Pt 1):6617–23. PubMed PMID: 18006762. 29. Marijnen CA, Nagtegaal ID, Klein Kranenbarg E, et al. No downstaging after short-term preoperative radiotherapy in rectal cancer patients. J Clin Oncol 2001 Apr 1;19(7):1976–84. PubMed PMID: 11283130. 30. Wolthuis AM, Penninckx F, Haustermans K, et al. Impact of interval between neoadjuvant chemoradiotherapy and TME for locally advanced rectal cancer on pathologic response and oncologic outcome. Ann Surg Oncol 2012 Sep;19(9):2833–41. PubMed PMID: 22451236. 31. Evans J, Tait D, Swift I, et al. Timing of surgery following preoperative therapy in rectal cancer: the need for a prospective randomized trial? Dis Colon Rectum 2011 Oct;54(10):1251–9. PubMed PMID: 21904139. 32. Shihab OC, Brown G, Daniels IR, Heald RJ, Quirke P, Moran BJ. Patients with low rectal cancer treated by abdominoperineal excision have worse tumors and higher involved margin rates compared with patients treated by anterior resection. Dis Colon Rectum 2010 Jan; 53(1):53–6. PubMed PMID: 20010351. 33. Holm T, Ljung A, Haggmark T, Jurell G, Lagergren J. Extended abdominoperineal resection with gluteus maximus flap reconstruction of the pelvic floor for rectal cancer. Br J Surg 2007 Feb;94(2):232– 8. PubMed PMID: 17143848. 34. Martling A, Cedermark B, Johansson H, Rutqvist LE, Holm T. The surgeon as a prognostic factor after the introduction of total mesorectal excision in the treatment of rectal cancer. Br J Surg 2002 Aug;89(8): 1008–13. PubMed PMID: 12153626. 35. Tekkis PP, Heriot AG, Smith J, et al. Comparison of circumferential margin involvement between restorative and nonrestorative resections for rectal cancer. Colorectal Dis 2005 Jul;7(4):369–74. PubMed PMID: 15932561. 36. Green BL, Marshall HC, Collinson F, et al. Long-term follow-up of the Medical Research Council CLASICC trial of conventional versus
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015
8
J. Homan et al. / EJSO xx (2015) 1e8
laparoscopically assisted resection in colorectal cancer. Br J Surg 2013 Jan;100(1):75–82. PubMed PMID: 23132548. 37. Ng SS, Leung KL, Lee JF, et al. Laparoscopic-assisted versus open abdominoperineal resection for low rectal cancer: a prospective randomized trial. Ann Surg Oncol 2008 Sep;15(9):2418–25. PubMed PMID: 18392659. 38. van der Pas MH, Haglind E, Cuesta MA, et al. Laparoscopic versus open surgery for rectal cancer (COLOR II): short-term outcomes of a randomised, phase 3 trial. Lancet Oncol 2013 Mar;14(3):210–8. PubMed PMID: 23395398.
39. Bonjer HJ, Deijen CL, Abis GA, et al. A randomized trial of laparoscopic versus open surgery for rectal cancer. N Engl J Med 2015 Apr 2;372(14):1324–32. PubMed PMID: 25830422. 40. Kulkarni T, Gollins S, Maw A, Hobson P, Byrne R, Widdowson D. Magnetic resonance imaging in rectal cancer downstaged using neoadjuvant chemoradiation: accuracy of prediction of tumour stage and circumferential resection margin status. Colorectal Dis 2008 Jun;10(5):479–89. PubMed PMID: 18318754.
Please cite this article in press as: Homan J, et al., Variation in circumferential resection margin: Reporting and involvement in the South-Netherlands, Eur J Surg Oncol (2015), http://dx.doi.org/10.1016/j.ejso.2015.07.015