Since the introduction of the TNM residual tumor (R) classification, the involvement of resection margins has been defined either as a microscopic (R1) or a macroscopic (R2) demonstration of tumor directly at the resection margin (“tumor transected”).
The recognition of the importance of the circumferential resection margin (CRM) in patients with rectal cancer patients raises the need for an alternative definition of resection margin involvement, namely, the importance of delineating tumor with a minimal distance from the CRM of ≤1 mm (CRM-positive) from tumor directly at the resection margin. The different use of both definitions of resection margin involvement prevents valid comparisons between reports on treatment results.
To avoid confusion by different definitions, the authors proposed including the minimal distance between tumor and resection margin into the current R classification.
In 1977, the American Joint Committee on Cancer (AJCC) recommended a residual tumor or R classification.1 Recognizing its outstanding prognostic importance, the R classification was adopted into the fourth edition of the TNM Classification of Malignant Tumors by the International Union Against Cancer (UICC)2 and in the corresponding third edition of the AJCC Manual for Staging of Cancer,3 and it has been retained up to the current TNM editions.4, 5
One of the basic elements of the R classification after surgical tumor removal is the assessment of the resection margins. Since the start of the R classification in 1977, a resection margin has been considered involved when there is macroscopic or microscopic tumor directly at the resection margin. However, in recent years, an alternative definition of resection margin involvement has gained increasing acceptance, at first for the circumferential resection margin (CRM) in rectal cancer6, 7 but also for other resection margins and other tumor entities. In addition, the R classification is designed to include both the local resection margin as well as distant metastasis (when present, often coded as R2).
In this report, both the traditional and alternative definitions of tumor involvement of resection margins are compared. The resulting problems, in particular the danger of invalid comparisons between reports on treatment results, are discussed. A proposal is presented for the integration of both definitions for further evaluation and uniform use of the R classification.
MATERIALS AND METHODS
Traditional Assessment of Resection Margin by R Classification
The AJCC 1977 residual tumor (or R) classification was applicable to patients who underwent definitive surgery and initially referred only to the site of the primary tumor.1 Recognizing the prognostic importance of the R classification after colorectal cancer surgery, it was propagated in 1980 by Hermanek et al8 and in 1981 by Newland et al.9 The current R classification was published 1987 as part of the TNM classification system2, 3 and has been retained up to the current TNM editions.4, 5 It defines tumor involvement of resection margins as macroscopic or microscopic demonstration of tumor directly at the resection margin (“tumor transected”).1, 3, 5, 8-13
Residual tumor includes not only continuous but also discontinuous extension of the primary tumor (“satellites,” “tumor deposits”), lymphatic and venous invasion (tumor cells in the lumen with adherence to the vessel wall and/or vessel wall invasion), perineural invasion, and lymph node metastasis.6, 7, 14, 15 In addition, if distant metastases are present, then these also are coded as either R1 (microscopic) or R2 (macroscopic).
Alternative Definition of Involved Resection Margins
With respect to the CRM (lateral, radial) in rectal cancer, an alternative definition of tumor involvement has evolved from data that demonstrated the relation between clinical outcome and the closest approach of the tumor to the CRM.6, 7, 16, 17 These data strongly support the following definitions: 1) CRM positive, tumor directly at the CRM or a minimal distance between the tumor and the CRM of ≤1 mm; and 2) CRM negative, a minimal distance between the tumor and the CRM >1 mm.
This “new” definition of tumor involvement of CRM has gained general acceptance in the United Kingdom. Currently, it is used in many other European and Asian countries18, 19 and is supported increasingly in the United States,20-22 but it has been declined in Australia.23 In the United Kingdom, the new definition also has been applied to the assessment of tumors at other sites.24 Corresponding data are available for esophageal carcinoma and adenocarcinomas of the pancreatic head, ampulla of Vater, and distal bile duct.25-29
Nagtegaal et al30 proposed a 2-mm cutoff between a positive CRM and a negative CRM. However, this proposal was based on data obtained after a median follow-up of 35 months and on 2-year local recurrence rates only. These findings could not be confirmed for patients from Leeds, who had substantially longer follow-up and 5-year local recurrence rates.31 According to a review of the literature by Glynne-Jones et al in 2006,32 the great majority of studies that dealt with CRM status used the ≤1 mm definition for positive CRM (91.1%; 7373 of 8094 patients). This definition also is used in more recently publications from the Netherlands.19, 33 In this context, it should be mentioned that the preoperative prediction by magnetic resonance imaging (MRI) for a CRM ≤1 mm has been established as more accurate than for a CRM ≤2 mm.34
RESULTS AND DISCUSSION
The assessment of resection margins can depend on fixation, gross dissection, marking of the margin, and sampling for histologic examination. For the CRM in rectal carcinomas, standards have been set.7, 35
In addition, the processing of tissue blocks for microscopy can influence results. If initial sections reveal tumor very close to the resection margin (≤1 mm), then, in some institutions, additional serial sections are prepared, which may demonstrate tumor directly at the margin. Thus, at least in studies, the histologic methods should be standardized and described in detail.
Differences, Advantages, and Disadvantages of the 2 Systems
The categories CRM negative and CRM positive are not directly comparable to the categories R0 and R1. CRM negative cannot be equated with R0, because R also includes the other resection margins and the presence of distant metastases. The same is valid for CRM positive and R1. In fact, there are 3 categories: 2 are combined in the R classification, and another 2 are combined in the CRM classification (Table 1).
Table 1. Circumferential Resection Margin Status and Residual (R) Tumor Classification
In rectal carcinoma, the prognosis, in particular the local recurrence rate, differs significantly for the 3 situations and for R1 versus R0 and CRM positive versus CRM negative (Fig. 1). For other tumor sites, data comparing the R classification with the CRM status are not available.
There is no doubt that, in rectal carcinoma, a subdivision of the relation of the tumor to the CRM into 3 categories, as proposed in Table 1, provides more detailed information on prognosis than the R0/R1 categories. This can be observed in the Leeds study,31 in which 165 of 586 patients (28.2%) had positive CRM status, but only 66 of 584 patients (11.3%) had margins that were classified as R1 (Fig. 1). Thus, for the identification of patients with a high risk of local recurrence, which would be an indication for adjuvant treatment, the CRM status is more useful. This is valid for neoadjuvant treatment, too; because, today, the preoperative prediction of CRM status by high-resolution MRI is possible with great reliability.18, 36, 37
Currently, the histologic finding of CRM positive is considered the most sensitive predictor of local recurrence and also is important for the prediction of distant metastasis and survival.19 Thus, the histopathologic CRM status increasingly is proposed as an intermediate endpoint (surrogate endpoint, early alternative endpoint) in studies.31, 34, 38-40
In summary, R classification and CRM status are different but may be complementary. The advantages of the CRM status are its sensitivity for prognosis, especially local recurrence, and its correlation with MRI in rectal cancer. It has been assessed only in these situations but may be important in other situations that have yet to be investigated, eg, in esophageal cancer and adenocarcinoma of the pancreatic head, ampulla of Vater, and distal bile duct. The advantages of the R classification are that it is applicable to all tumor sites and that it summarizes the presence of residual disease in the patient. Therefore, neither the traditional R categories R0 and R1 nor the categories CRM positive and CRM negative are completely satisfactory for all tumor sites. In contrast to the continued use of separate classifications, implementation of a single system with combined categories has advantages.
Danger of Different Definitions of Tumor-Involved Resection Margins
Because both the traditional R classification and the CRM status have advantages, both systems are used, but their definitions are not always uniform. In some centers, R1 and R0 do not denote the traditional categories of the R classification but, rather, the alternative (CRM) definitions.
Treatment results from an institution presented only in relation to the traditional R classification do not allow a reliable comparison with the treatment results from another institution using only the alternative (CRM) definition. An example can be observed in the data from Figure 1. If an institution presents its data only for CRM-negative patients, then the crude local recurrence rate is 10% (42 of 421 patients), which is considerably lower than the rate for R0 patients (69 of 518 patients; 13.3%—a relative reduction of 25%!). Conversely, the rate of curative resections, defined according to the traditional R0 category, is higher than for the alternative definition (CRM negative): 88.7% versus 71.8%.31 Thus, some confusion arises in the interpretation of data. A uniformly defined nomenclature for the assessment of resection margins is needed,32 and some authors already have attempted it.
The minimal distance between tumor and resection margin of ≤1 mm has been proposed by a new category: “R0 close”.41 this has been used by the German tumor documentation system since 200242 and has been recommended by the Professional Association of German Pathologists and the German Society of Pathology since 2007.43 In the same sense “R ≤1 mm” is used by Borschitz and Junginger.44
The Austrian Breast and Colorectal Cancer Study Group45 proposed using both the R classification and the CRM status and differentiating between “R (UICC)” and “R (Quirke).” Such an overlapping double classification poses an inherent risk of confusion and inconsistent application. To guarantee the possibility of comparisons between treatment results from different institutions as well as the use of both definitions, an expansion of the traditional R classification by integration of the alternative definition of the CRM status seems attractive.
Proposal to Expand the Current R Classification
During the last 10 to 15 years, multiple studies in patients with rectal carcinoma have demonstrated that outcome is related not only to the histologic demonstration of tumor directly at the resection margin but also to the minimal distance between tumor and resection margin.19 In particular, outcome is significantly adversely affected by surgical clearance of ≤1 mm versus >1 mm. The consideration of direct tumor involvement and of the minimal distance of tumor from the resection margin also influences the indication for neoadjuvant and adjuvant treatment. Thus, we propose including the minimal distance between tumor and resection margin into the current R classification. Accordingly, instead of R0 and R1, 3 new categories are introduced (Table 2). This classification is in accordance with the recent proposal of Dent et al23 for separate documentation of direct tumor involvement of CRM as well as minimal distance between tumor and CRM and was incorporated into the United Kingdom Royal College of Pathologists minimum dataset.24
Tumor indicates not only continuous extension but also discontinuous extension of the primary tumor (“satellites,” “tumor deposits”), lymphatic and venous invasion (tumor cells in the lumen with adherence to the vessel wall and/or vessel wall invasion), perineural invasion, and lymph node metastasis.
R0 through R2c are changes in classification.
R0 > 1 mm
No residual tumor, minimal distance between tumor and resection margin; margin ≥1 mm†
R1 ≤ 1 mm
No residual tumor, minimal distance between tumor and resection margin; margin ≤1 mm†
Microscopic residual tumor, tumor directly at the resection margin (tumor transected)†
By adopting this proposal, the current confusion caused by the use of different definitions of tumor involvement of resection margins can be eliminated. Currently, it is not possible to compare treatment results (survival, local recurrence rates) presented with respect to CRM status (sometimes described as R0/R1) with those using the traditional R0/R1 categories. By using the proposed categories of expanded R classifications, it is possible to assign the traditional R classification status and simultaneously to classify cases according to the alternative definition of tumor involvement of resection margins (CRM status) (see Table 1). Therefore, a comparison of new data with former publications will be possible.
In addition, R2 is used frequently, although incorrectly, for local status alone. To avoid confusion, we propose a subdivision for this category as well, with R2a for local macroscopic residual tumor and R2b for distant macroscopic residual tumor and with R2c for residual tumor in both local and distant sites.
The value of the new proposed classification for patients with rectal carcinoma is obvious, as demonstrated above. It seems likely that this may be valid for other tumor entities, but that remains to be proven. The proposed system may facilitate such validity studies.
Conflict of Interest Disclosures
Phil Quirke receives support for colorectal cancer research from a Yorkshire Cancer Research Program Grant.
Leslie H. Sobin is a United States government employee.