Prognostic value of the circumferential resection margin and its definitions in esophageal cancer patients after neoadjuvant chemoradiotherapy

Prognostic value of the circumferential resection margin and its definitions in esophageal cancer... SUMMARY The accepted importance of a positive circumferential resection margin (CRM) (defined as R1 in the TNM classification) is based on histopathology of the resection specimen obtained after primary surgery in esophageal cancer patients. The aim of this study is to look for the prognostic value of CRM after neoadjuvant chemoradiotherapy and to compare the clinical significance of a histologically CRM < 1 mm from the cut margin (Royal College of Pathologists definition of R1) to a positive cut margin (College of American Pathologists definition of R1) and to ≥1 mm margin (R0) resections in patients with ypT3-esophageal tumors after neoadjuvant chemoradiotherapy.  Between 2000 and 2014, 458 patients who received esophagectomy after neoadjuvant chemoradiation therapy were selected. Overall (OS) and disease-free survival (DFS) were calculated by means of Kaplan–Meier curves and compared by Cox regression analysis.  There were 163 (35.9%) patients who had a ypT3 tumor; in 118 (72.4%) resection was complete (R0). In 37 (22.7%) patients a CRM < 1 mm was found and 8 (4.9%) had a circumferential R1-resection. CRM involvement was inversely correlated with tumor regression grading, lymph node capsular involvement, and number of positive lymph nodes.  On univariate analysis, no statistically significant difference was found between R0-resection and CRM < 1 mm (P = 0.103) for OS, but DFS showed a significant difference (P = 0.025). Circumferential R1-resections showed a significant difference compared to R0-resections for OS and DFS (both P = 0.002). In multivariate analysis, extracapsular lymph node involvement and circumferential R1-resection were withheld as independent prognosticators for OS, whereas extracapsular lymph node involvement, absence of regression on the primary tumor and circumferential R1-resection were withheld for DFS. After correcting for different variables in the multivariate model, CRM < 1 mm showed no statistical difference compared to R0-resections neither for OS nor for DFS.  After neoadjuvant chemoradiotherapy, CRM is correlated with biological behavior of the tumor and with therapy response. Furthermore it is an independent prognosticator for OS and DFS. However CRM < 1 mm itself is no independent prognosticator for OS nor DFS survival in multivariable analysis. These results suggest that the definition of R1-resection should be limited to true invasion of the section plane. INTRODUCTION Surgical extirpation—and more specific: complete resection—is the gold standard in the curative treatment of resectable esophageal carcinoma.1,2 In the majority of cases, proximal and distal negative resection margins can be achieved, but due to the anatomical relationship of the esophagus with other mediastinal structures, obtaining a microscopically negative circumferential resection margin (CRM) can be challenging with reported positive resection margins (proximal, distal, and circumferential) in 8.6% to 25% in esophagectomy series.3,4 Moreover, two different definitions of microscopic complete resection are utilized: the College of American Pathologists (CAP) considers a positive CRM as tumor found at the cut margin of resection,5 whereas the Royal College of Pathologists (RCP) defines a positive CRM as tumor cells found within 1 mm of the cut margins.6 It is a matter of debate which definition should be utilized for prognostic purposes.7–9 Since recent randomized trials10,11 showed a better local control and survival after neoadjuvant treatment, locally advanced esophageal tumors are nowadays treated with neoadjuvant therapy followed by surgery in order to improve the rate of negative circumferential resection margins (CRM) and survival. The aim of this study is twofold: first to look for the prognostic value of CRM involvement after neoadjuvant chemoradiotherapy on survival. Second to compare the prognostic significance of the R1 definitions by respectively CAP (positive at the cut margin) and RCP (<1 mm from the cut margin) to a R0 resection defined as a CRM ≥ 1 mm. MATERIALS AND METHODS Patients Between 2000 and 2014, all patients with carcinoma of the esophagus or the gastroesophageal junction, treated with neoadjuvant chemoradiation therapy followed by surgery, were retrospectively extracted from our prospectively maintained database. All patients with a residual tumor beyond the muscular layer (= ypT3 according to the TNM 7th edn.) were withheld. Exclusion criteria were histology other than squamous cell carcinoma (SCC) or adenocarcinoma (AC), macroscopic positive surgical margin (R2-esophagectomy), and microscopic positive proximal or distal surgical margin. Neoadjuvant treatment The most frequently used chemotherapy scheme (86%) in both SCC patients and AC patients was two cycles of cisplatin 80 mg/m² on days 1 and 21 and 5-fluorouracil 800 mg/m² on days 1–4 and 21–24, combined with a total radiation dose varying between 40 and 45 Gy in fractions of 1.8 Gy. In 14% of patients adapted schemes were used. Surgery Esophagectomy was performed by open technique (91%) using a left thoracoabdominal approach or a combination of right posterolateral thoracotomy and laparotomy both with cervical or intrathoracic anastomosis. In some patients (9%), a minimal invasive resection was performed using a right thoracoscopy, laparoscopy, and cervical anastomosis. Lymph node dissection in all patients consisted of at least a standard two-field lymphadenectomy according to the definitions of the consensus meeting of the International Society for the Diseases of the Esophagus 1995.12 Pathologic examination All resection specimens were evaluated and prospectively recorded by a senior gastrointestinal pathologist. Standard HE stained slides were prepared from each paraffin block for microscopic examination and additional HEs or keratin stains were performed were needed. Three groups were defined according to the circumferential resection margin: group 1 (R0): tumor-free resection margin of 1 mm or more, group 2 (CRM < 1 mm): tumor cells within 1 mm from the surgical margin and group 3 (R1): tumor cells found at the cut margin. Mandard tumor regression grade (TRG) was classified according to Mandard's tumor regression scheme.13 Extracapsular extension of tumor in the lymph nodes was defined as the presence of tumor cells that outgrow the lymph node and extent beyond the capsule of the lymph node, but in continuity with the tumor mass in the nodal parenchyma. Follow up Patients were evaluated every 3 months during the first year, every 6 months thereafter until the 6th year, after which patients were followed on a yearly basis. Follow-up investigations included clinical examination and serum biochemistry with carcinoembryonal antigen level. When indicated, flexible upper GI endoscopy was performed. Imaging studies (CT and increasingly 18-fluorodeoxyglucose position emission tomography over the last 10 years) were obtained every 6 months and on yearly base after 3 years. Statistical analysis Age, gender, and tumor characteristics including AJCC 7th edn. pathologic stage, histology, grade, R-status, Mandard TRG, number of invaded lymph nodes and extra- (EC-LNI) versus intra- (IC-LNI) capsular lymph node invasion were recorded. These values were compared with the use of t-test or Chi square and overall survival (OS) and disease-free survival (DFS) were estimated by means of the Kaplan–Meier curves and compared by Log-rank tests. Univariable and multivariable analysis were performed by Cox proportional hazard regression analysis. The multivariable model was constructed by means of a stratified Cox regression analysis, using the variables found significant in the univariate analysis: variables with a P < 0.05 were included in a stepwise Cox regression model. P-values smaller than 0.05 are considered significant. All analyses were performed using the IBM SPSS Statistics software, version 23. RESULTS During the study period, a total of 1610 patients underwent esophagectomy for cancer. Out of 458 patients treated with neoadjuvant chemoradiation therapy followed by surgery, 163 were withheld with a tumor classified as ypT3 (Fig. 1). Group 1 (tumor-free resection margin of ≥1 mm) consisted of 118 patients (72.4%), group 2 (residual tumor cells within 1 mm) consisted of 37 patients (22.7%), and group 3 (tumor cells found at the cut margin) of 8 patients (4.9%). Mean follow up of these patients was 42,3 (32,6- 52,0 95% CI) months. Characteristics of the three subgroups are described and compared in Table 1. Statistically significant differences are seen between the subgroups concerning Mandard tumor regression grading (P = 0.005), incidence of lymph node involvement (P = 0.001), and mean number of invaded lymph nodes (P < 0.0001). Histology distribution was not different between the three groups. Fig. 1 View largeDownload slide Flow chart of the study population. ADC, adenocarcinoma; nCRT, neoadjuvant chemoradiation therapy; nCT, neoadjuvant chemotherapy; NT, neoadjuvant therapy; SCC, squamous cell carcinoma. Fig. 1 View largeDownload slide Flow chart of the study population. ADC, adenocarcinoma; nCRT, neoadjuvant chemoradiation therapy; nCT, neoadjuvant chemotherapy; NT, neoadjuvant therapy; SCC, squamous cell carcinoma. Table 1 Clinicopathological characteristics of the three groups according to circumferential resection margin       Group 1  Group 2  Group 3        All  (R0)  (CRM < 1 mm)  (R1 lateral)  Significance      n  %  n  %  n  %  n  %  (χ²)  n =    163  100%  118  72%  37  23%  8  5%  p =  Gender  Female  33  20%  24  20%  7  19%  2  25%      0.926    Male  130  80%  94  80%  30  81%  6  75%    Mean age (range)    61  [37–80]  57  [37–80]  54  [37–74]  62  [43–73]  0.782  Histology  ADC  118  72%  87  74%  27  73%  4  50%      0.347    SCC  45  28%  31  26%  10  27%  4  50%    Mandard  Grade 2  57  35%  46  39%  11  30%          Grade 3  54  33%  42  36%  9  24%  3  38%      0.005    Grade 4  47  29%  29  25%  13  35%  5  63%      Grade 5  5  3%  1  1%  4  11%        LN-status  LN−  61  37%  55  47%  4  11%  2  25%      LN + IC  48  29%  32  27%  15  41%  1  13%  0.001    LN + EC  54  33%  31  26%  18  49%  5  63%    Mean number positive LN’s    2,6  [1–27]  2,0  [1–15]  3,7  [1–27]  5,1  [1–12]   <0.0001        Group 1  Group 2  Group 3        All  (R0)  (CRM < 1 mm)  (R1 lateral)  Significance      n  %  n  %  n  %  n  %  (χ²)  n =    163  100%  118  72%  37  23%  8  5%  p =  Gender  Female  33  20%  24  20%  7  19%  2  25%      0.926    Male  130  80%  94  80%  30  81%  6  75%    Mean age (range)    61  [37–80]  57  [37–80]  54  [37–74]  62  [43–73]  0.782  Histology  ADC  118  72%  87  74%  27  73%  4  50%      0.347    SCC  45  28%  31  26%  10  27%  4  50%    Mandard  Grade 2  57  35%  46  39%  11  30%          Grade 3  54  33%  42  36%  9  24%  3  38%      0.005    Grade 4  47  29%  29  25%  13  35%  5  63%      Grade 5  5  3%  1  1%  4  11%        LN-status  LN−  61  37%  55  47%  4  11%  2  25%      LN + IC  48  29%  32  27%  15  41%  1  13%  0.001    LN + EC  54  33%  31  26%  18  49%  5  63%    Mean number positive LN’s    2,6  [1–27]  2,0  [1–15]  3,7  [1–27]  5,1  [1–12]   <0.0001  ADC, adenocarcinoma; EC, extracapsular involvement; IC, intracapsular involvement; LN, lymph node, SCC, squamous cell carcinoma. View Large In group 1, 15 patients (12.7%) had locoregional recurrence, 37 patients (31.4%) had distant metastatic disease, 12 patients (10.2%) had both types of recurrences and 54 patients (45.8%) had no recurrence. In group 2, 6 patients (16.2%) had locoregional recurrence, 14 patients (37.8%) had distant metastatic disease, 6patients (16.2%) had both types of recurrences and 11 patients (29.7%) had no recurrence. In group 3, there were no isolated locoregional recurrences, 1 patient (12.5%) had isolated distant metastatic disease, 2 patients (25.0%) had both types of recurrences, and 5 patients (62.5%) had no recurrence. Recurrence patterns were not significantly different between groups. Median OS for group 1, group 2, and group 3 is 23.2, 15.9, and 3.5 months respectively. Overall 3-year survival and 5-year survival were 36.9% and 23.4%, respectively, for group 1, 12.5% and 0.0%, respectively, for group 2 and 17.9% and 3.4%, respectively, for group 3. Median DFS for group 1, group 2, and group 3 is 12.4, 7.3, and 3.5 months, respectively. Disease-free 3-year survival and 5-year survival were 26.1% and 20.0%, respectively, for group 1, 9.4% and 9.4% respectively for group 2 and 0.0% and 0.0% respectively for group 3. OS and DFS of Group 1 compared to group 3 were significantly different (both P = 0.002). OS between group 1 and group 2 was not significantly different (P = 0.103) but DFS however reached significance (P = 0.025). Comparison of OS and DFS between group 2 and group 3 showed no significant difference for OS (P = 0.061) nor for DFS (P = 0.075). Kaplan–Meier survival curves of the three subgroups for OS and DFS are shown in Figures 2 and 3, respectively. Fig. 2 View largeDownload slide Overall survival according to the microscopic resection margin. R0 = tumor-free resection margin of 1 mm or more; CRM < 1 mm = tumor cells within 1 mm from the surgical margin; R1 = tumor cells found at the cut margin. Fig. 2 View largeDownload slide Overall survival according to the microscopic resection margin. R0 = tumor-free resection margin of 1 mm or more; CRM < 1 mm = tumor cells within 1 mm from the surgical margin; R1 = tumor cells found at the cut margin. Fig. 3 View largeDownload slide Disease-free survival according to the microscopic resection margin. R0 = tumor-free resection margin of 1 mm or more; CRM < 1 mm = tumor cells within 1 mm from the surgical margin; R1 = tumor cells found at the cut margin. Fig. 3 View largeDownload slide Disease-free survival according to the microscopic resection margin. R0 = tumor-free resection margin of 1 mm or more; CRM < 1 mm = tumor cells within 1 mm from the surgical margin; R1 = tumor cells found at the cut margin. Subanalysis per histology showed a median overall survival for group 1, group 2, and group 3 of 23.2, 16.3, and 6.4 months, respectively in AC patients. Median DFS in AC patients for group 1, group 2, and group 3 is 12.5, 7.7, and 5.0 months, respectively. OS of Group 1 compared to group 3 was significantly different (P = 0.020). DFS of Group 1 compared to group 3 showed a nonsignificant trend (P = 0.058). OS between group 1 and group 2 was not significantly different (P = 0.213), just as DFS was not significantly different (P = 0.053). Comparison of OS and DFS between group 2 and group 3 showed no significant difference for OS (P = 0.069) nor for DFS (P = 0.389). Median OS for group 1, group 2, and group 3 of 23.7, 15.5, and 0.6 months, respectively, in SCC patients. Median DFS in SCC patients for group 1, group 2, and group 3 is 12.0, 6.2, and 0.6 months, respectively. OS and DFS of Group 1 compared to group 3 were significantly different (P = 0.038 and P = 0.021, respectively). OS and DFS between group 1 and group 2 were not significantly different (P = 0.321 and P = 0.257, respectively). Comparison of OS and DFS between group 2 and group 3 showed no significant difference for OS (P = 0.399) nor for DFS survival (P = 0.203). The univariable analysis for overall survival and disease-free survival for the whole group is shown in Table 2. Circumferential resection margin, number of positive lymph nodes, and extent of lymph node invasion were prognosticators for OS and DFS. Mandard TRG was only a prognosticator for DFS. Table 2 Univariable analysis OS and DFS   OS  DFS    Sig.  HR  95.0% CI for HR  Sig.  HR  95.0% CI for HR        Lower  Upper      Lower  Upper  Gender  0.717  0.92  0.57  1.47  0.511  0.86  0.55  1.35  Age  0.151  1.01  1.00  1.03  0.364  1.01  0.99  1.02  Histology; SCC = ref  0.457  0.86  0.58  1.28  0.325  0.83  0.57  1.21  Mandard TRG; TRG2 = ref  0.071         <0.0001         TRG 3  0.12  1.44  0.91  2.28  0.01  1.78  1.15  2.75   TRG 4  0.103  1.49  0.92  2.40  0.076  1.51  0.96  2.37   TRG5  0.018  3.55  1.24  10.16   <0.0001  7.50  2.80  20.08  LNStatus; pN0 = ref  0.012         <0.0001         IC  0.193  1.37  0.85  2.19  0.069  1.51  0.97  2.34   EC  0.003  1.96  1.26  3.06   <0.0001  2.41  1.58  3.66  Num POS LN’s  0.034  1.05  1.00  1.10  0.004  1.06  1.02  1.10  CRM < 1 mm = ref  0.01        0.002         R0  0.11  0.70  0.45  1.09  0.027  0.63  0.41  0.95   R1  0.056  2.16  0.98  4.78  0.088  1.97  0.90  4.31    OS  DFS    Sig.  HR  95.0% CI for HR  Sig.  HR  95.0% CI for HR        Lower  Upper      Lower  Upper  Gender  0.717  0.92  0.57  1.47  0.511  0.86  0.55  1.35  Age  0.151  1.01  1.00  1.03  0.364  1.01  0.99  1.02  Histology; SCC = ref  0.457  0.86  0.58  1.28  0.325  0.83  0.57  1.21  Mandard TRG; TRG2 = ref  0.071         <0.0001         TRG 3  0.12  1.44  0.91  2.28  0.01  1.78  1.15  2.75   TRG 4  0.103  1.49  0.92  2.40  0.076  1.51  0.96  2.37   TRG5  0.018  3.55  1.24  10.16   <0.0001  7.50  2.80  20.08  LNStatus; pN0 = ref  0.012         <0.0001         IC  0.193  1.37  0.85  2.19  0.069  1.51  0.97  2.34   EC  0.003  1.96  1.26  3.06   <0.0001  2.41  1.58  3.66  Num POS LN’s  0.034  1.05  1.00  1.10  0.004  1.06  1.02  1.10  CRM < 1 mm = ref  0.01        0.002         R0  0.11  0.70  0.45  1.09  0.027  0.63  0.41  0.95   R1  0.056  2.16  0.98  4.78  0.088  1.97  0.90  4.31  EC, extracapsular involvement; IC, intracapsular involvement; LN, lymph node; SCC, squamous cell carcinoma; TRG, tumor regression grade. View Large In multivariable analysis, EC-LNI and circumferential R1-resection (CAP definition) were withheld as independent prognosticators for OS (HR 1.83 [1.15–2.91] and HR 2.86 [1.37–6.01] respectively), whereas EC-LNI, Mandard TRG 5, and circumferential R1-resection were withheld for DFS (HR 1.99 [1.28–3.10], HR 5.24 [1.88–14.61], and HR 2.47 [1.17–5.22] respectively) (Table 3). Table 3 Multivariable analysis OS and DFS   OS  DFS    Sig.  HR  95.0% CI for HR  Sig.  HR  95.0% CI for HR        Lower  Upper      Lower  Upper  R-status (R0 = ref)  0.020        0.050         CRM < 1 mm  0.454  1.19  0.75  1.90  0.313  1.27  0.80  2.00   R1 Lateral  0.005  2.86  1.37  6.01  0.018  2.47  1.17  5.22  LN-Status (pN0 = ref)  0.038        0.009         IC-LNI  0.202  1.37  0.85  2.22  0.205  1.34  0.85  2.12   EC-LNI  0.011  1.83  1.15  2.91  0.002  1.99  1.28  3.10  Mandard (TRG 2 = ref)  0.290        0.009         TRG 3  0.777        0.056  1.56  0.99  2.45   TRG 4  0.629        0.245  1.32  0.83  2.09   TRG 5  0.104        0.002  5.24  1.88  14.61  Age  0.060        0.071        Gender  0.950        0.911        Histology  0.450        0.245        Number of POSITIVE LN  0.854        0.774          OS  DFS    Sig.  HR  95.0% CI for HR  Sig.  HR  95.0% CI for HR        Lower  Upper      Lower  Upper  R-status (R0 = ref)  0.020        0.050         CRM < 1 mm  0.454  1.19  0.75  1.90  0.313  1.27  0.80  2.00   R1 Lateral  0.005  2.86  1.37  6.01  0.018  2.47  1.17  5.22  LN-Status (pN0 = ref)  0.038        0.009         IC-LNI  0.202  1.37  0.85  2.22  0.205  1.34  0.85  2.12   EC-LNI  0.011  1.83  1.15  2.91  0.002  1.99  1.28  3.10  Mandard (TRG 2 = ref)  0.290        0.009         TRG 3  0.777        0.056  1.56  0.99  2.45   TRG 4  0.629        0.245  1.32  0.83  2.09   TRG 5  0.104        0.002  5.24  1.88  14.61  Age  0.060        0.071        Gender  0.950        0.911        Histology  0.450        0.245        Number of POSITIVE LN  0.854        0.774        EC, extracapsular involvement; IC, intracapsular involvement; LN, lymph node; TRG, tumor regression grade. View Large CRM < 1 mm itself, showed no statistical difference compared to R0-resections neither for OS (HR 1.19 [0.75–1.90], P = 0.454) nor for DFS (HR 1.27 [0.80–2.00], P = 0.313). Because the RCP definition of a positive margin (CRM < 1 mm) includes also and by definition the CAP definition (CRM = tumor cells at the cut margin) survival between CAP and RCP criteria was also compared. Median OS of R0 according to the CAP definition (18.5 months) was significantly different from median OS of R1 according to the CAP definition (3.5 months) (P = 0.003) and median OS of R0 according to the RCP definition (23.2 months) was significantly different from median OS of R1 according to the RCP definition (15.5 months) (P = 0.016). There was no significant difference in median OS between R0 according to the CAP definition and R0 according to the RCP definition (P = 0.583) nor in median OS between R1 according to the CAP definition and R1 according to the RCP definition (P = 0.135). Median DFS of R0 according to the CAP definition (10.7 months) was significantly different from median OS of R1 according to the CAP definition (3.5 months) (P = 0.004) and median DFS of R0 according to the RCP definition (12.4 months) was significantly different from median OS of R1 according to the RCP definition (6.6 months) (P = 0.003). There was no significant difference in median DFS between R0 according to the CAP definition and R0 according to the RCP definition (P = 0.467) nor in median DFS between R1 according to the CAP definition and R1 according to the RCP definition (P = 0.147). DISCUSSION This study shows that CRM in an esophagectomy specimen after neoadjuvant chemoradiation is an independent prognosticator for OS and DFS. CRM < 1 mm however, is not an independent prognostic factor for OS nor for DFS. CRM involvement is also correlated with biologic tumor behavior and the response on neoadjuvant therapy, independently of histology. CRM as a prognostic factor was originally described in esophagectomy specimens after primary surgery.14 Discussion about the definition of CRM based on different pathological definitions (CAP or RCP)5,6 led to contradictory results in primary surgery patients and mixed patient groups with different types of multimodal therapy.15–24 In so far that very recently even a cutoff margin at 500 μm has been proposed as an additional criterium to define the extent of CRM involvement.25 Two meta-analyses dealing with the CRM issue after primary surgery7,8 found that a positive CRM was associated with poor prognosis, both for CAP and RCP definitions. Unfortunately it appears that the impact of CRM involvement often is estimated on a population consisting of T1 to T3 tumors creating thus a bias, because CRM involvement of a T1 or T2 tumor is a surgical failure and should not be mixed up with the T3 tumors. Another limitation is that in many studies a subgroup analysis according lymph node status is lacking. For the latter also information on the extent of lymphadenectomy will be required but is mostly lacking. In patients undergoing neoadjuvant chemoradiation therapy, Chao et al.26 showed that in 151 ypT3 esophageal squamous cell carcinoma patients after neoadjuvant chemoradiation the incidence of local recurrence was significantly higher in group R1 and in the group CRM < 1 mm as compared to the group CRM ≥ 1 mm (P = 0.05 and P = 0.04, respectively). Furthermore, the 5 year DFS differed significantly between CRM ≥ 1 mm (40%), CRM < 1 mm (23%) and R1 according to CAP (6%) (all p < 0.001), leading them to conclude that the CRM distance could bear additional prognostic value. On the other hand, the major drawback of the previous series is the use of univariable analysis methods only possibly missing confounders like lymph node involvement and response to neoadjuvant treatment. In their retrospective study using multivariate and propensity matched analysis, Harvin et al.27 could not find any difference in local recurrence nor in overall survival between positive and negative CRM according to RCP criteria in 160 ypT3 esophageal adenocarcinoma patients after neoadjuvant chemoradiation suggesting that CRM < 1 mm (R1 RCP) is closer to CRM ≥ 1 mm (R0) and as such does not appear to be of any additional value. These results confirm our findings. Indeed, in our study on ypT3 tumors after neoadjuvant chemoradiation, we could not find any added value of using the RCP criteria in multivariable analysis. Furthermore, Markar et al.3 showed that the incidence of positive CRM according to the CAP criteria was correlated with Mandard TRG and number of positive lymph nodes. Nevertheless in the multivariable analysis they found that in patients receiving neoadjuvant therapy a positive CRM according to the CAP criteria remained an independent predictor for poor prognosis. The current series confirms these findings, showing a correlation between involved CRM and significantly higher Mandard TRG and number of positive lymph nodes. Moreover the study also indicated that higher extracapsular tumoral spread in lymph nodes in patients is increasing with more narrowly CRM. Chao et al.26 also found a higher rate of distant recurrences in patients with R1-resection according to the CAP criteria, where they only expected more local recurrences. Although DFS and OS of only 3.5 months in our series suggest a similar problem, i.e., more distant recurrences in group 3 (R1) suggesting a more aggressive biological behavior in this subset of patients, we were not able to prove this because not all patients in this small group died from oncological reasons. In our opinion, the presence or absence of a positive CRM in an esophagectomy specimen after neoadjuvant chemoradiation therapy is seemingly clearly correlated with a more aggressive biologic tumor behavior leading to poorer survival and is inversely related with tumor response to neoadjuvant therapy. This is also visually depicted in the survival curves of the three CRM groups for OS and DFS, with a clear trend toward worse survival in patients with increasingly narrowly CRM although they do not always reach statistical significance, probably because our series is underpowered in group 3 (R1). Direct comparison of survival—OS as well as DFS—between R0 according to the CAP definition and R0 according to the RCP definition, showed no statistical differences in our series, nor did it for comparison of survival between respective R1 definitions. This is probably because of small series and R1 definition according to RCP criteria is including R1 definition according to CAP criteria and R0 definition according to CAP criteria is including R0 definition according to RCP criteria. Therefore these criteria cannot correctly be used in multivariable analysis. On the other hand, after correction for different confounders in the multivariable model, CRM < 1 mm itself is not different from CRM ≥ 1 mm in our study with 72% adenocarcinoma patients. Therefore one should be careful with interpreting CRM < 1 mm as a negative prognostic factor for survival, as this seemingly worse survival is probably more dictated by other criteria like lymph node invasion and tumor regression grading, as shown in our multivariable analysis and supported by Markar et al.3 Therefore we suggest using the CAP criteria for defining CRM in esophagectomy specimens after neoadjuvant radiochemotherapy. A limitation of this study is obviously its retrospective character, although all parameters were prospectively recorded in our database by a senior GI pathologist. Furthermore, the relatively small numbers in groups 2 and 3 make statistical interpretation somewhat difficult due to the lack of statistical power. However, to our knowledge, there are no larger series about CRM in ypT3 resection specimens in patients with esophageal carcinoma after neoadjuvant chemoradiation therapy than present study and the two described studies by Chao et al.26 and Harvin et al.27 Analysis of pooled data of these studies could probably enhance the power to answer our questions about the significance of CRM after neoadjuvant chemoradiation in esophageal carcinoma in the future. We conclude that CRM in esophageal cancer patients after neoadjuvant chemoradiation therapy, irrespective the histology, is an important prognostic factor, reflecting the tumor response on neoadjuvant therapy. CRM < 1 mm, after correction for different variables in the multivariate analysis, is not different from R0-resections ≥1 mm neither for OS nor for DFS. These results suggest that the definition of R1-resection should be limited to true invasion of the section plane. Notes Specific author contributions: Study design, analysis and interpretation of data: Lieven Depypere, Toni Lerut, Philippe Nafteux; drafting of the manuscript: Lieven Depypere, Johnny Moons, Toni Lerut, Philippe Nafteux; critical revision: Lieven Depypere, Johnny Moons, Toni Lerut, Gert De Hertogh, Chloë Peters, Xavier Sagaert, Willy Coosemans, Hans Van Veer, Philippe Nafteux; study design, acquisition of data, analysis and interpretation of data: Johnny Moons; acquisition of data: Gert De Hertogh, Chloë Peters, Xavier Sagaert, Willy Coosemans, Hans Van Veer. ACKNOWLEDGMENTS All authors declare to have no potential conflicts of interest and they have no other disclosures to declare. References 1 Lordick F, Mariette C, Haustermans K, Obermannová R, Arnold D. Oesophageal cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol  2016; 27: v50– 7. Google Scholar CrossRef Search ADS PubMed  2 Ajani J A, D’Amico T A, Almhanna K. National comprehensive cancer network. Esophageal and esophagogastric junction cancers, version 1.2015. J Natl Compr Canc Netw  2015; 13: 194– 227. Google Scholar CrossRef Search ADS PubMed  3 Markar S R, Gronnier C, Duhamel A et al.   Significance of microscopically incomplete resection margin after esophagectomy for esophageal cancer. Ann Surg  2016; 263: 712– 8. Google Scholar CrossRef Search ADS PubMed  4 Gilbert S, Martel A B, Seely A J et al.   Prognostic significance of a positive radial margin after esophageal cancer resection. J Thorac Cardiovasc Surg  2015; 149: 548– 55. Google Scholar CrossRef Search ADS PubMed  5 Mapstone N. Dataset for the Histopathological Reporting of Oesophageal Carcinoma , 2nd edn. London: Royal College of Pathologists, 2006. 6 College of American Pathologists. Surgical Pathology Cancer Case Summary (Checklist): Esophagus . College of American Pathologists: Northfield, 2005. 7 Wu J, Chen Q, Teng L et al.   Prognostic significance of positive circumferential resection margin in esophageal cancer: a systematic review and meta-analysis. Ann Thorac Surg  2014; 97: 446– 53. Google Scholar CrossRef Search ADS PubMed  8 Chan D, Reid T, Howell I, Lewis W. Systematic review and meta-analysis of the influence of circumferential resection margin involvement on survival in patients with operable oesophageal cancer. Br J Surg  2013; 100: 456– 64. Google Scholar CrossRef Search ADS PubMed  9 O’Neill J, Stephens N, Save V et al.   Defining a positive circumferential resection margin in oesophageal cancer and its implications for adjuvant treatment. Br J Surg  2013; 100: 1055– 63. Google Scholar CrossRef Search ADS PubMed  10 Shapiro J, van Lanschot J J, Hulshof M C et al.   Neoadjuvant chemoradiotherapy plus surgery versus surgery alone for oesophageal or junctional cancer (CROSS): long-term results of a randomised controlled trial. Lancet Oncol  2015; 16: 1090– 8. Google Scholar CrossRef Search ADS PubMed  11 Kelsen D P, Winter K A, Gunderson L L et al.   Long-term results of RTOG trial 891 (USA Intergroup 113): a random assignment trial comparison of chemotherapy followed by surgery compared with surgery alone for esophageal cancer. J Clin Oncol  2007; 25: 3719– 25. Google Scholar CrossRef Search ADS PubMed  12 Fumagalli U. Resective surgery for cancer of the thoracic esophagus. Results of a consensus conference held at the vith world congress of the international society for diseases of the esophagus. Dis Esophagus  1996; 9: 30– 8. 13 Mandard A M, Dalibard F, Mandard J C et al.   Pathologic assessment of tumor regression after preoperative chemoradiotherapy of esophageal carcinoma. Clinicopathologic Correlations. Cancer  1994; 73: 2680– 6. 14 Sagar P M, Johnston D, McMahon M J, Dixon M F, Quirke P. Significance of circumferential resection margin involvement after oesophagectomy for cancer. Br J Surg  1993; 80: 1386– 8. Google Scholar CrossRef Search ADS PubMed  15 Gilbert S, Martel A, Seely A. Prognostic significance of a positive radial margin after esophageal cancer resection. J Thorac Cardiovasc Surg  2015; 149: 548– 55. Google Scholar CrossRef Search ADS PubMed  16 Verhage R, Zandvoort H, ten Kate F, van Hillegersberg R. How to define a positive circumferential resection margin in T3 adenocarcinoma of the esophagus? Am J Surg Pathol  2011; 35: 919– 26. Google Scholar CrossRef Search ADS PubMed  17 Hulshoff J, Faiz Z, Karrenbeld A et al.   Prognostic value of circumferential resection margin in esophageal cancer patients after neoadjuvant chemoradiotherapy. Ann Surg Oncol  2015; 22: S1301– 9. Google Scholar CrossRef Search ADS PubMed  18 Okada N, Fujii S, Fujita T et al.   The prognostic significance of the positive circumferential resection margin in pathologic T3 squamous cell carcinoma of the esophagus with or without neoadjuvant chemotherapy. Surgery  2016; 159: 441– 50. Google Scholar CrossRef Search ADS PubMed  19 Deeter M, Dorer R, Kuppusamy M et al.   Assessment of criteria and clinical significance of circumferential resection margins in esophageal cancer. Arch Surg  2009; 144: 618– 24. Google Scholar CrossRef Search ADS PubMed  20 Rao V, Yeung M, Cooke J et al.   Comparison of circumferential resection margin clearance criteria with survival after surgery for cancer of the esophagus. J Surg Oncol  2012; 105: 745– 9. Google Scholar CrossRef Search ADS PubMed  21 Reid T, Chan D, Roberts S et al.   Prognostic significance of circumferential resection margin involvement following esophagectomy for cancer and the predicitive role of endoluminal ultrasonography. Br J Cancer  2012; 107: 1925– 31. Google Scholar CrossRef Search ADS PubMed  22 Salih T, Jose P, Mehta S et al.   Prognostic significance of cancer within 1 mm of the circumferential resection margin in esophageal cancer patients following neo-adjuvant chemotherapy. Eur J Cardiothorac Surg  2013; 43: 562– 7. Google Scholar CrossRef Search ADS PubMed  23 O’Farrell N J, Donohoe C L, Muldoon C et al.   Lack of independent significance of a close (<1 mm) circumferential resection margin involvement in esophageal and junctional cancer. Ann Surg Oncol  2013; 20: 2727– 33. Google Scholar CrossRef Search ADS PubMed  24 Sillah K, Pritchard S, Watkins G et al.   The degree of circumferential tumour involvement as a prognostic factor in esophageal cancer. Eur J Cardiothorac Surg  2009; 36: 368– 73. Google Scholar CrossRef Search ADS PubMed  25 Lee G D, Lee S E, Kim K M et al.   New 3-tiered circumferential resection margin criteria in esophageal squamous cell carcinoma. Ann Surg  2015; 262: 965– 71. Google Scholar CrossRef Search ADS PubMed  26 Chao Y, Yeh C, Chang H et al.   Impact of circumferential resection margin distance on locoregional recurrence and survival after chemoradiotherapy in esophageal squamous cell carcinoma. Ann Surg Oncol  2011; 18: 529– 34. Google Scholar CrossRef Search ADS PubMed  27 Harvin J, Lahat G, Correa A et al.   Neoadjuvant chemoradiotherapy followed by surgery for esophageal adenocarcinoma: significance of microscopically positive circumferential radial margins. J Thorac Cardiovasc Surg  2012; 143: 412– 20. Google Scholar CrossRef Search ADS PubMed  © The Authors 2017. Published by Oxford University Press on behalf of International Society for Diseases of the Esophagus. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Diseases of the Esophagus Oxford University Press

Prognostic value of the circumferential resection margin and its definitions in esophageal cancer patients after neoadjuvant chemoradiotherapy

Loading next page...
 
/lp/ou_press/prognostic-value-of-the-circumferential-resection-margin-and-its-9PUkQWE0mz
Publisher
The International Society for Diseases of the Esophagus
Copyright
© The Authors 2017. Published by Oxford University Press on behalf of International Society for Diseases of the Esophagus. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
1120-8694
eISSN
1442-2050
D.O.I.
10.1093/dote/dox117
Publisher site
See Article on Publisher Site

Abstract

SUMMARY The accepted importance of a positive circumferential resection margin (CRM) (defined as R1 in the TNM classification) is based on histopathology of the resection specimen obtained after primary surgery in esophageal cancer patients. The aim of this study is to look for the prognostic value of CRM after neoadjuvant chemoradiotherapy and to compare the clinical significance of a histologically CRM < 1 mm from the cut margin (Royal College of Pathologists definition of R1) to a positive cut margin (College of American Pathologists definition of R1) and to ≥1 mm margin (R0) resections in patients with ypT3-esophageal tumors after neoadjuvant chemoradiotherapy.  Between 2000 and 2014, 458 patients who received esophagectomy after neoadjuvant chemoradiation therapy were selected. Overall (OS) and disease-free survival (DFS) were calculated by means of Kaplan–Meier curves and compared by Cox regression analysis.  There were 163 (35.9%) patients who had a ypT3 tumor; in 118 (72.4%) resection was complete (R0). In 37 (22.7%) patients a CRM < 1 mm was found and 8 (4.9%) had a circumferential R1-resection. CRM involvement was inversely correlated with tumor regression grading, lymph node capsular involvement, and number of positive lymph nodes.  On univariate analysis, no statistically significant difference was found between R0-resection and CRM < 1 mm (P = 0.103) for OS, but DFS showed a significant difference (P = 0.025). Circumferential R1-resections showed a significant difference compared to R0-resections for OS and DFS (both P = 0.002). In multivariate analysis, extracapsular lymph node involvement and circumferential R1-resection were withheld as independent prognosticators for OS, whereas extracapsular lymph node involvement, absence of regression on the primary tumor and circumferential R1-resection were withheld for DFS. After correcting for different variables in the multivariate model, CRM < 1 mm showed no statistical difference compared to R0-resections neither for OS nor for DFS.  After neoadjuvant chemoradiotherapy, CRM is correlated with biological behavior of the tumor and with therapy response. Furthermore it is an independent prognosticator for OS and DFS. However CRM < 1 mm itself is no independent prognosticator for OS nor DFS survival in multivariable analysis. These results suggest that the definition of R1-resection should be limited to true invasion of the section plane. INTRODUCTION Surgical extirpation—and more specific: complete resection—is the gold standard in the curative treatment of resectable esophageal carcinoma.1,2 In the majority of cases, proximal and distal negative resection margins can be achieved, but due to the anatomical relationship of the esophagus with other mediastinal structures, obtaining a microscopically negative circumferential resection margin (CRM) can be challenging with reported positive resection margins (proximal, distal, and circumferential) in 8.6% to 25% in esophagectomy series.3,4 Moreover, two different definitions of microscopic complete resection are utilized: the College of American Pathologists (CAP) considers a positive CRM as tumor found at the cut margin of resection,5 whereas the Royal College of Pathologists (RCP) defines a positive CRM as tumor cells found within 1 mm of the cut margins.6 It is a matter of debate which definition should be utilized for prognostic purposes.7–9 Since recent randomized trials10,11 showed a better local control and survival after neoadjuvant treatment, locally advanced esophageal tumors are nowadays treated with neoadjuvant therapy followed by surgery in order to improve the rate of negative circumferential resection margins (CRM) and survival. The aim of this study is twofold: first to look for the prognostic value of CRM involvement after neoadjuvant chemoradiotherapy on survival. Second to compare the prognostic significance of the R1 definitions by respectively CAP (positive at the cut margin) and RCP (<1 mm from the cut margin) to a R0 resection defined as a CRM ≥ 1 mm. MATERIALS AND METHODS Patients Between 2000 and 2014, all patients with carcinoma of the esophagus or the gastroesophageal junction, treated with neoadjuvant chemoradiation therapy followed by surgery, were retrospectively extracted from our prospectively maintained database. All patients with a residual tumor beyond the muscular layer (= ypT3 according to the TNM 7th edn.) were withheld. Exclusion criteria were histology other than squamous cell carcinoma (SCC) or adenocarcinoma (AC), macroscopic positive surgical margin (R2-esophagectomy), and microscopic positive proximal or distal surgical margin. Neoadjuvant treatment The most frequently used chemotherapy scheme (86%) in both SCC patients and AC patients was two cycles of cisplatin 80 mg/m² on days 1 and 21 and 5-fluorouracil 800 mg/m² on days 1–4 and 21–24, combined with a total radiation dose varying between 40 and 45 Gy in fractions of 1.8 Gy. In 14% of patients adapted schemes were used. Surgery Esophagectomy was performed by open technique (91%) using a left thoracoabdominal approach or a combination of right posterolateral thoracotomy and laparotomy both with cervical or intrathoracic anastomosis. In some patients (9%), a minimal invasive resection was performed using a right thoracoscopy, laparoscopy, and cervical anastomosis. Lymph node dissection in all patients consisted of at least a standard two-field lymphadenectomy according to the definitions of the consensus meeting of the International Society for the Diseases of the Esophagus 1995.12 Pathologic examination All resection specimens were evaluated and prospectively recorded by a senior gastrointestinal pathologist. Standard HE stained slides were prepared from each paraffin block for microscopic examination and additional HEs or keratin stains were performed were needed. Three groups were defined according to the circumferential resection margin: group 1 (R0): tumor-free resection margin of 1 mm or more, group 2 (CRM < 1 mm): tumor cells within 1 mm from the surgical margin and group 3 (R1): tumor cells found at the cut margin. Mandard tumor regression grade (TRG) was classified according to Mandard's tumor regression scheme.13 Extracapsular extension of tumor in the lymph nodes was defined as the presence of tumor cells that outgrow the lymph node and extent beyond the capsule of the lymph node, but in continuity with the tumor mass in the nodal parenchyma. Follow up Patients were evaluated every 3 months during the first year, every 6 months thereafter until the 6th year, after which patients were followed on a yearly basis. Follow-up investigations included clinical examination and serum biochemistry with carcinoembryonal antigen level. When indicated, flexible upper GI endoscopy was performed. Imaging studies (CT and increasingly 18-fluorodeoxyglucose position emission tomography over the last 10 years) were obtained every 6 months and on yearly base after 3 years. Statistical analysis Age, gender, and tumor characteristics including AJCC 7th edn. pathologic stage, histology, grade, R-status, Mandard TRG, number of invaded lymph nodes and extra- (EC-LNI) versus intra- (IC-LNI) capsular lymph node invasion were recorded. These values were compared with the use of t-test or Chi square and overall survival (OS) and disease-free survival (DFS) were estimated by means of the Kaplan–Meier curves and compared by Log-rank tests. Univariable and multivariable analysis were performed by Cox proportional hazard regression analysis. The multivariable model was constructed by means of a stratified Cox regression analysis, using the variables found significant in the univariate analysis: variables with a P < 0.05 were included in a stepwise Cox regression model. P-values smaller than 0.05 are considered significant. All analyses were performed using the IBM SPSS Statistics software, version 23. RESULTS During the study period, a total of 1610 patients underwent esophagectomy for cancer. Out of 458 patients treated with neoadjuvant chemoradiation therapy followed by surgery, 163 were withheld with a tumor classified as ypT3 (Fig. 1). Group 1 (tumor-free resection margin of ≥1 mm) consisted of 118 patients (72.4%), group 2 (residual tumor cells within 1 mm) consisted of 37 patients (22.7%), and group 3 (tumor cells found at the cut margin) of 8 patients (4.9%). Mean follow up of these patients was 42,3 (32,6- 52,0 95% CI) months. Characteristics of the three subgroups are described and compared in Table 1. Statistically significant differences are seen between the subgroups concerning Mandard tumor regression grading (P = 0.005), incidence of lymph node involvement (P = 0.001), and mean number of invaded lymph nodes (P < 0.0001). Histology distribution was not different between the three groups. Fig. 1 View largeDownload slide Flow chart of the study population. ADC, adenocarcinoma; nCRT, neoadjuvant chemoradiation therapy; nCT, neoadjuvant chemotherapy; NT, neoadjuvant therapy; SCC, squamous cell carcinoma. Fig. 1 View largeDownload slide Flow chart of the study population. ADC, adenocarcinoma; nCRT, neoadjuvant chemoradiation therapy; nCT, neoadjuvant chemotherapy; NT, neoadjuvant therapy; SCC, squamous cell carcinoma. Table 1 Clinicopathological characteristics of the three groups according to circumferential resection margin       Group 1  Group 2  Group 3        All  (R0)  (CRM < 1 mm)  (R1 lateral)  Significance      n  %  n  %  n  %  n  %  (χ²)  n =    163  100%  118  72%  37  23%  8  5%  p =  Gender  Female  33  20%  24  20%  7  19%  2  25%      0.926    Male  130  80%  94  80%  30  81%  6  75%    Mean age (range)    61  [37–80]  57  [37–80]  54  [37–74]  62  [43–73]  0.782  Histology  ADC  118  72%  87  74%  27  73%  4  50%      0.347    SCC  45  28%  31  26%  10  27%  4  50%    Mandard  Grade 2  57  35%  46  39%  11  30%          Grade 3  54  33%  42  36%  9  24%  3  38%      0.005    Grade 4  47  29%  29  25%  13  35%  5  63%      Grade 5  5  3%  1  1%  4  11%        LN-status  LN−  61  37%  55  47%  4  11%  2  25%      LN + IC  48  29%  32  27%  15  41%  1  13%  0.001    LN + EC  54  33%  31  26%  18  49%  5  63%    Mean number positive LN’s    2,6  [1–27]  2,0  [1–15]  3,7  [1–27]  5,1  [1–12]   <0.0001        Group 1  Group 2  Group 3        All  (R0)  (CRM < 1 mm)  (R1 lateral)  Significance      n  %  n  %  n  %  n  %  (χ²)  n =    163  100%  118  72%  37  23%  8  5%  p =  Gender  Female  33  20%  24  20%  7  19%  2  25%      0.926    Male  130  80%  94  80%  30  81%  6  75%    Mean age (range)    61  [37–80]  57  [37–80]  54  [37–74]  62  [43–73]  0.782  Histology  ADC  118  72%  87  74%  27  73%  4  50%      0.347    SCC  45  28%  31  26%  10  27%  4  50%    Mandard  Grade 2  57  35%  46  39%  11  30%          Grade 3  54  33%  42  36%  9  24%  3  38%      0.005    Grade 4  47  29%  29  25%  13  35%  5  63%      Grade 5  5  3%  1  1%  4  11%        LN-status  LN−  61  37%  55  47%  4  11%  2  25%      LN + IC  48  29%  32  27%  15  41%  1  13%  0.001    LN + EC  54  33%  31  26%  18  49%  5  63%    Mean number positive LN’s    2,6  [1–27]  2,0  [1–15]  3,7  [1–27]  5,1  [1–12]   <0.0001  ADC, adenocarcinoma; EC, extracapsular involvement; IC, intracapsular involvement; LN, lymph node, SCC, squamous cell carcinoma. View Large In group 1, 15 patients (12.7%) had locoregional recurrence, 37 patients (31.4%) had distant metastatic disease, 12 patients (10.2%) had both types of recurrences and 54 patients (45.8%) had no recurrence. In group 2, 6 patients (16.2%) had locoregional recurrence, 14 patients (37.8%) had distant metastatic disease, 6patients (16.2%) had both types of recurrences and 11 patients (29.7%) had no recurrence. In group 3, there were no isolated locoregional recurrences, 1 patient (12.5%) had isolated distant metastatic disease, 2 patients (25.0%) had both types of recurrences, and 5 patients (62.5%) had no recurrence. Recurrence patterns were not significantly different between groups. Median OS for group 1, group 2, and group 3 is 23.2, 15.9, and 3.5 months respectively. Overall 3-year survival and 5-year survival were 36.9% and 23.4%, respectively, for group 1, 12.5% and 0.0%, respectively, for group 2 and 17.9% and 3.4%, respectively, for group 3. Median DFS for group 1, group 2, and group 3 is 12.4, 7.3, and 3.5 months, respectively. Disease-free 3-year survival and 5-year survival were 26.1% and 20.0%, respectively, for group 1, 9.4% and 9.4% respectively for group 2 and 0.0% and 0.0% respectively for group 3. OS and DFS of Group 1 compared to group 3 were significantly different (both P = 0.002). OS between group 1 and group 2 was not significantly different (P = 0.103) but DFS however reached significance (P = 0.025). Comparison of OS and DFS between group 2 and group 3 showed no significant difference for OS (P = 0.061) nor for DFS (P = 0.075). Kaplan–Meier survival curves of the three subgroups for OS and DFS are shown in Figures 2 and 3, respectively. Fig. 2 View largeDownload slide Overall survival according to the microscopic resection margin. R0 = tumor-free resection margin of 1 mm or more; CRM < 1 mm = tumor cells within 1 mm from the surgical margin; R1 = tumor cells found at the cut margin. Fig. 2 View largeDownload slide Overall survival according to the microscopic resection margin. R0 = tumor-free resection margin of 1 mm or more; CRM < 1 mm = tumor cells within 1 mm from the surgical margin; R1 = tumor cells found at the cut margin. Fig. 3 View largeDownload slide Disease-free survival according to the microscopic resection margin. R0 = tumor-free resection margin of 1 mm or more; CRM < 1 mm = tumor cells within 1 mm from the surgical margin; R1 = tumor cells found at the cut margin. Fig. 3 View largeDownload slide Disease-free survival according to the microscopic resection margin. R0 = tumor-free resection margin of 1 mm or more; CRM < 1 mm = tumor cells within 1 mm from the surgical margin; R1 = tumor cells found at the cut margin. Subanalysis per histology showed a median overall survival for group 1, group 2, and group 3 of 23.2, 16.3, and 6.4 months, respectively in AC patients. Median DFS in AC patients for group 1, group 2, and group 3 is 12.5, 7.7, and 5.0 months, respectively. OS of Group 1 compared to group 3 was significantly different (P = 0.020). DFS of Group 1 compared to group 3 showed a nonsignificant trend (P = 0.058). OS between group 1 and group 2 was not significantly different (P = 0.213), just as DFS was not significantly different (P = 0.053). Comparison of OS and DFS between group 2 and group 3 showed no significant difference for OS (P = 0.069) nor for DFS (P = 0.389). Median OS for group 1, group 2, and group 3 of 23.7, 15.5, and 0.6 months, respectively, in SCC patients. Median DFS in SCC patients for group 1, group 2, and group 3 is 12.0, 6.2, and 0.6 months, respectively. OS and DFS of Group 1 compared to group 3 were significantly different (P = 0.038 and P = 0.021, respectively). OS and DFS between group 1 and group 2 were not significantly different (P = 0.321 and P = 0.257, respectively). Comparison of OS and DFS between group 2 and group 3 showed no significant difference for OS (P = 0.399) nor for DFS survival (P = 0.203). The univariable analysis for overall survival and disease-free survival for the whole group is shown in Table 2. Circumferential resection margin, number of positive lymph nodes, and extent of lymph node invasion were prognosticators for OS and DFS. Mandard TRG was only a prognosticator for DFS. Table 2 Univariable analysis OS and DFS   OS  DFS    Sig.  HR  95.0% CI for HR  Sig.  HR  95.0% CI for HR        Lower  Upper      Lower  Upper  Gender  0.717  0.92  0.57  1.47  0.511  0.86  0.55  1.35  Age  0.151  1.01  1.00  1.03  0.364  1.01  0.99  1.02  Histology; SCC = ref  0.457  0.86  0.58  1.28  0.325  0.83  0.57  1.21  Mandard TRG; TRG2 = ref  0.071         <0.0001         TRG 3  0.12  1.44  0.91  2.28  0.01  1.78  1.15  2.75   TRG 4  0.103  1.49  0.92  2.40  0.076  1.51  0.96  2.37   TRG5  0.018  3.55  1.24  10.16   <0.0001  7.50  2.80  20.08  LNStatus; pN0 = ref  0.012         <0.0001         IC  0.193  1.37  0.85  2.19  0.069  1.51  0.97  2.34   EC  0.003  1.96  1.26  3.06   <0.0001  2.41  1.58  3.66  Num POS LN’s  0.034  1.05  1.00  1.10  0.004  1.06  1.02  1.10  CRM < 1 mm = ref  0.01        0.002         R0  0.11  0.70  0.45  1.09  0.027  0.63  0.41  0.95   R1  0.056  2.16  0.98  4.78  0.088  1.97  0.90  4.31    OS  DFS    Sig.  HR  95.0% CI for HR  Sig.  HR  95.0% CI for HR        Lower  Upper      Lower  Upper  Gender  0.717  0.92  0.57  1.47  0.511  0.86  0.55  1.35  Age  0.151  1.01  1.00  1.03  0.364  1.01  0.99  1.02  Histology; SCC = ref  0.457  0.86  0.58  1.28  0.325  0.83  0.57  1.21  Mandard TRG; TRG2 = ref  0.071         <0.0001         TRG 3  0.12  1.44  0.91  2.28  0.01  1.78  1.15  2.75   TRG 4  0.103  1.49  0.92  2.40  0.076  1.51  0.96  2.37   TRG5  0.018  3.55  1.24  10.16   <0.0001  7.50  2.80  20.08  LNStatus; pN0 = ref  0.012         <0.0001         IC  0.193  1.37  0.85  2.19  0.069  1.51  0.97  2.34   EC  0.003  1.96  1.26  3.06   <0.0001  2.41  1.58  3.66  Num POS LN’s  0.034  1.05  1.00  1.10  0.004  1.06  1.02  1.10  CRM < 1 mm = ref  0.01        0.002         R0  0.11  0.70  0.45  1.09  0.027  0.63  0.41  0.95   R1  0.056  2.16  0.98  4.78  0.088  1.97  0.90  4.31  EC, extracapsular involvement; IC, intracapsular involvement; LN, lymph node; SCC, squamous cell carcinoma; TRG, tumor regression grade. View Large In multivariable analysis, EC-LNI and circumferential R1-resection (CAP definition) were withheld as independent prognosticators for OS (HR 1.83 [1.15–2.91] and HR 2.86 [1.37–6.01] respectively), whereas EC-LNI, Mandard TRG 5, and circumferential R1-resection were withheld for DFS (HR 1.99 [1.28–3.10], HR 5.24 [1.88–14.61], and HR 2.47 [1.17–5.22] respectively) (Table 3). Table 3 Multivariable analysis OS and DFS   OS  DFS    Sig.  HR  95.0% CI for HR  Sig.  HR  95.0% CI for HR        Lower  Upper      Lower  Upper  R-status (R0 = ref)  0.020        0.050         CRM < 1 mm  0.454  1.19  0.75  1.90  0.313  1.27  0.80  2.00   R1 Lateral  0.005  2.86  1.37  6.01  0.018  2.47  1.17  5.22  LN-Status (pN0 = ref)  0.038        0.009         IC-LNI  0.202  1.37  0.85  2.22  0.205  1.34  0.85  2.12   EC-LNI  0.011  1.83  1.15  2.91  0.002  1.99  1.28  3.10  Mandard (TRG 2 = ref)  0.290        0.009         TRG 3  0.777        0.056  1.56  0.99  2.45   TRG 4  0.629        0.245  1.32  0.83  2.09   TRG 5  0.104        0.002  5.24  1.88  14.61  Age  0.060        0.071        Gender  0.950        0.911        Histology  0.450        0.245        Number of POSITIVE LN  0.854        0.774          OS  DFS    Sig.  HR  95.0% CI for HR  Sig.  HR  95.0% CI for HR        Lower  Upper      Lower  Upper  R-status (R0 = ref)  0.020        0.050         CRM < 1 mm  0.454  1.19  0.75  1.90  0.313  1.27  0.80  2.00   R1 Lateral  0.005  2.86  1.37  6.01  0.018  2.47  1.17  5.22  LN-Status (pN0 = ref)  0.038        0.009         IC-LNI  0.202  1.37  0.85  2.22  0.205  1.34  0.85  2.12   EC-LNI  0.011  1.83  1.15  2.91  0.002  1.99  1.28  3.10  Mandard (TRG 2 = ref)  0.290        0.009         TRG 3  0.777        0.056  1.56  0.99  2.45   TRG 4  0.629        0.245  1.32  0.83  2.09   TRG 5  0.104        0.002  5.24  1.88  14.61  Age  0.060        0.071        Gender  0.950        0.911        Histology  0.450        0.245        Number of POSITIVE LN  0.854        0.774        EC, extracapsular involvement; IC, intracapsular involvement; LN, lymph node; TRG, tumor regression grade. View Large CRM < 1 mm itself, showed no statistical difference compared to R0-resections neither for OS (HR 1.19 [0.75–1.90], P = 0.454) nor for DFS (HR 1.27 [0.80–2.00], P = 0.313). Because the RCP definition of a positive margin (CRM < 1 mm) includes also and by definition the CAP definition (CRM = tumor cells at the cut margin) survival between CAP and RCP criteria was also compared. Median OS of R0 according to the CAP definition (18.5 months) was significantly different from median OS of R1 according to the CAP definition (3.5 months) (P = 0.003) and median OS of R0 according to the RCP definition (23.2 months) was significantly different from median OS of R1 according to the RCP definition (15.5 months) (P = 0.016). There was no significant difference in median OS between R0 according to the CAP definition and R0 according to the RCP definition (P = 0.583) nor in median OS between R1 according to the CAP definition and R1 according to the RCP definition (P = 0.135). Median DFS of R0 according to the CAP definition (10.7 months) was significantly different from median OS of R1 according to the CAP definition (3.5 months) (P = 0.004) and median DFS of R0 according to the RCP definition (12.4 months) was significantly different from median OS of R1 according to the RCP definition (6.6 months) (P = 0.003). There was no significant difference in median DFS between R0 according to the CAP definition and R0 according to the RCP definition (P = 0.467) nor in median DFS between R1 according to the CAP definition and R1 according to the RCP definition (P = 0.147). DISCUSSION This study shows that CRM in an esophagectomy specimen after neoadjuvant chemoradiation is an independent prognosticator for OS and DFS. CRM < 1 mm however, is not an independent prognostic factor for OS nor for DFS. CRM involvement is also correlated with biologic tumor behavior and the response on neoadjuvant therapy, independently of histology. CRM as a prognostic factor was originally described in esophagectomy specimens after primary surgery.14 Discussion about the definition of CRM based on different pathological definitions (CAP or RCP)5,6 led to contradictory results in primary surgery patients and mixed patient groups with different types of multimodal therapy.15–24 In so far that very recently even a cutoff margin at 500 μm has been proposed as an additional criterium to define the extent of CRM involvement.25 Two meta-analyses dealing with the CRM issue after primary surgery7,8 found that a positive CRM was associated with poor prognosis, both for CAP and RCP definitions. Unfortunately it appears that the impact of CRM involvement often is estimated on a population consisting of T1 to T3 tumors creating thus a bias, because CRM involvement of a T1 or T2 tumor is a surgical failure and should not be mixed up with the T3 tumors. Another limitation is that in many studies a subgroup analysis according lymph node status is lacking. For the latter also information on the extent of lymphadenectomy will be required but is mostly lacking. In patients undergoing neoadjuvant chemoradiation therapy, Chao et al.26 showed that in 151 ypT3 esophageal squamous cell carcinoma patients after neoadjuvant chemoradiation the incidence of local recurrence was significantly higher in group R1 and in the group CRM < 1 mm as compared to the group CRM ≥ 1 mm (P = 0.05 and P = 0.04, respectively). Furthermore, the 5 year DFS differed significantly between CRM ≥ 1 mm (40%), CRM < 1 mm (23%) and R1 according to CAP (6%) (all p < 0.001), leading them to conclude that the CRM distance could bear additional prognostic value. On the other hand, the major drawback of the previous series is the use of univariable analysis methods only possibly missing confounders like lymph node involvement and response to neoadjuvant treatment. In their retrospective study using multivariate and propensity matched analysis, Harvin et al.27 could not find any difference in local recurrence nor in overall survival between positive and negative CRM according to RCP criteria in 160 ypT3 esophageal adenocarcinoma patients after neoadjuvant chemoradiation suggesting that CRM < 1 mm (R1 RCP) is closer to CRM ≥ 1 mm (R0) and as such does not appear to be of any additional value. These results confirm our findings. Indeed, in our study on ypT3 tumors after neoadjuvant chemoradiation, we could not find any added value of using the RCP criteria in multivariable analysis. Furthermore, Markar et al.3 showed that the incidence of positive CRM according to the CAP criteria was correlated with Mandard TRG and number of positive lymph nodes. Nevertheless in the multivariable analysis they found that in patients receiving neoadjuvant therapy a positive CRM according to the CAP criteria remained an independent predictor for poor prognosis. The current series confirms these findings, showing a correlation between involved CRM and significantly higher Mandard TRG and number of positive lymph nodes. Moreover the study also indicated that higher extracapsular tumoral spread in lymph nodes in patients is increasing with more narrowly CRM. Chao et al.26 also found a higher rate of distant recurrences in patients with R1-resection according to the CAP criteria, where they only expected more local recurrences. Although DFS and OS of only 3.5 months in our series suggest a similar problem, i.e., more distant recurrences in group 3 (R1) suggesting a more aggressive biological behavior in this subset of patients, we were not able to prove this because not all patients in this small group died from oncological reasons. In our opinion, the presence or absence of a positive CRM in an esophagectomy specimen after neoadjuvant chemoradiation therapy is seemingly clearly correlated with a more aggressive biologic tumor behavior leading to poorer survival and is inversely related with tumor response to neoadjuvant therapy. This is also visually depicted in the survival curves of the three CRM groups for OS and DFS, with a clear trend toward worse survival in patients with increasingly narrowly CRM although they do not always reach statistical significance, probably because our series is underpowered in group 3 (R1). Direct comparison of survival—OS as well as DFS—between R0 according to the CAP definition and R0 according to the RCP definition, showed no statistical differences in our series, nor did it for comparison of survival between respective R1 definitions. This is probably because of small series and R1 definition according to RCP criteria is including R1 definition according to CAP criteria and R0 definition according to CAP criteria is including R0 definition according to RCP criteria. Therefore these criteria cannot correctly be used in multivariable analysis. On the other hand, after correction for different confounders in the multivariable model, CRM < 1 mm itself is not different from CRM ≥ 1 mm in our study with 72% adenocarcinoma patients. Therefore one should be careful with interpreting CRM < 1 mm as a negative prognostic factor for survival, as this seemingly worse survival is probably more dictated by other criteria like lymph node invasion and tumor regression grading, as shown in our multivariable analysis and supported by Markar et al.3 Therefore we suggest using the CAP criteria for defining CRM in esophagectomy specimens after neoadjuvant radiochemotherapy. A limitation of this study is obviously its retrospective character, although all parameters were prospectively recorded in our database by a senior GI pathologist. Furthermore, the relatively small numbers in groups 2 and 3 make statistical interpretation somewhat difficult due to the lack of statistical power. However, to our knowledge, there are no larger series about CRM in ypT3 resection specimens in patients with esophageal carcinoma after neoadjuvant chemoradiation therapy than present study and the two described studies by Chao et al.26 and Harvin et al.27 Analysis of pooled data of these studies could probably enhance the power to answer our questions about the significance of CRM after neoadjuvant chemoradiation in esophageal carcinoma in the future. We conclude that CRM in esophageal cancer patients after neoadjuvant chemoradiation therapy, irrespective the histology, is an important prognostic factor, reflecting the tumor response on neoadjuvant therapy. CRM < 1 mm, after correction for different variables in the multivariate analysis, is not different from R0-resections ≥1 mm neither for OS nor for DFS. These results suggest that the definition of R1-resection should be limited to true invasion of the section plane. Notes Specific author contributions: Study design, analysis and interpretation of data: Lieven Depypere, Toni Lerut, Philippe Nafteux; drafting of the manuscript: Lieven Depypere, Johnny Moons, Toni Lerut, Philippe Nafteux; critical revision: Lieven Depypere, Johnny Moons, Toni Lerut, Gert De Hertogh, Chloë Peters, Xavier Sagaert, Willy Coosemans, Hans Van Veer, Philippe Nafteux; study design, acquisition of data, analysis and interpretation of data: Johnny Moons; acquisition of data: Gert De Hertogh, Chloë Peters, Xavier Sagaert, Willy Coosemans, Hans Van Veer. ACKNOWLEDGMENTS All authors declare to have no potential conflicts of interest and they have no other disclosures to declare. References 1 Lordick F, Mariette C, Haustermans K, Obermannová R, Arnold D. Oesophageal cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol  2016; 27: v50– 7. Google Scholar CrossRef Search ADS PubMed  2 Ajani J A, D’Amico T A, Almhanna K. National comprehensive cancer network. Esophageal and esophagogastric junction cancers, version 1.2015. J Natl Compr Canc Netw  2015; 13: 194– 227. Google Scholar CrossRef Search ADS PubMed  3 Markar S R, Gronnier C, Duhamel A et al.   Significance of microscopically incomplete resection margin after esophagectomy for esophageal cancer. Ann Surg  2016; 263: 712– 8. Google Scholar CrossRef Search ADS PubMed  4 Gilbert S, Martel A B, Seely A J et al.   Prognostic significance of a positive radial margin after esophageal cancer resection. J Thorac Cardiovasc Surg  2015; 149: 548– 55. Google Scholar CrossRef Search ADS PubMed  5 Mapstone N. Dataset for the Histopathological Reporting of Oesophageal Carcinoma , 2nd edn. London: Royal College of Pathologists, 2006. 6 College of American Pathologists. Surgical Pathology Cancer Case Summary (Checklist): Esophagus . College of American Pathologists: Northfield, 2005. 7 Wu J, Chen Q, Teng L et al.   Prognostic significance of positive circumferential resection margin in esophageal cancer: a systematic review and meta-analysis. Ann Thorac Surg  2014; 97: 446– 53. Google Scholar CrossRef Search ADS PubMed  8 Chan D, Reid T, Howell I, Lewis W. Systematic review and meta-analysis of the influence of circumferential resection margin involvement on survival in patients with operable oesophageal cancer. Br J Surg  2013; 100: 456– 64. Google Scholar CrossRef Search ADS PubMed  9 O’Neill J, Stephens N, Save V et al.   Defining a positive circumferential resection margin in oesophageal cancer and its implications for adjuvant treatment. Br J Surg  2013; 100: 1055– 63. Google Scholar CrossRef Search ADS PubMed  10 Shapiro J, van Lanschot J J, Hulshof M C et al.   Neoadjuvant chemoradiotherapy plus surgery versus surgery alone for oesophageal or junctional cancer (CROSS): long-term results of a randomised controlled trial. Lancet Oncol  2015; 16: 1090– 8. Google Scholar CrossRef Search ADS PubMed  11 Kelsen D P, Winter K A, Gunderson L L et al.   Long-term results of RTOG trial 891 (USA Intergroup 113): a random assignment trial comparison of chemotherapy followed by surgery compared with surgery alone for esophageal cancer. J Clin Oncol  2007; 25: 3719– 25. Google Scholar CrossRef Search ADS PubMed  12 Fumagalli U. Resective surgery for cancer of the thoracic esophagus. Results of a consensus conference held at the vith world congress of the international society for diseases of the esophagus. Dis Esophagus  1996; 9: 30– 8. 13 Mandard A M, Dalibard F, Mandard J C et al.   Pathologic assessment of tumor regression after preoperative chemoradiotherapy of esophageal carcinoma. Clinicopathologic Correlations. Cancer  1994; 73: 2680– 6. 14 Sagar P M, Johnston D, McMahon M J, Dixon M F, Quirke P. Significance of circumferential resection margin involvement after oesophagectomy for cancer. Br J Surg  1993; 80: 1386– 8. Google Scholar CrossRef Search ADS PubMed  15 Gilbert S, Martel A, Seely A. Prognostic significance of a positive radial margin after esophageal cancer resection. J Thorac Cardiovasc Surg  2015; 149: 548– 55. Google Scholar CrossRef Search ADS PubMed  16 Verhage R, Zandvoort H, ten Kate F, van Hillegersberg R. How to define a positive circumferential resection margin in T3 adenocarcinoma of the esophagus? Am J Surg Pathol  2011; 35: 919– 26. Google Scholar CrossRef Search ADS PubMed  17 Hulshoff J, Faiz Z, Karrenbeld A et al.   Prognostic value of circumferential resection margin in esophageal cancer patients after neoadjuvant chemoradiotherapy. Ann Surg Oncol  2015; 22: S1301– 9. Google Scholar CrossRef Search ADS PubMed  18 Okada N, Fujii S, Fujita T et al.   The prognostic significance of the positive circumferential resection margin in pathologic T3 squamous cell carcinoma of the esophagus with or without neoadjuvant chemotherapy. Surgery  2016; 159: 441– 50. Google Scholar CrossRef Search ADS PubMed  19 Deeter M, Dorer R, Kuppusamy M et al.   Assessment of criteria and clinical significance of circumferential resection margins in esophageal cancer. Arch Surg  2009; 144: 618– 24. Google Scholar CrossRef Search ADS PubMed  20 Rao V, Yeung M, Cooke J et al.   Comparison of circumferential resection margin clearance criteria with survival after surgery for cancer of the esophagus. J Surg Oncol  2012; 105: 745– 9. Google Scholar CrossRef Search ADS PubMed  21 Reid T, Chan D, Roberts S et al.   Prognostic significance of circumferential resection margin involvement following esophagectomy for cancer and the predicitive role of endoluminal ultrasonography. Br J Cancer  2012; 107: 1925– 31. Google Scholar CrossRef Search ADS PubMed  22 Salih T, Jose P, Mehta S et al.   Prognostic significance of cancer within 1 mm of the circumferential resection margin in esophageal cancer patients following neo-adjuvant chemotherapy. Eur J Cardiothorac Surg  2013; 43: 562– 7. Google Scholar CrossRef Search ADS PubMed  23 O’Farrell N J, Donohoe C L, Muldoon C et al.   Lack of independent significance of a close (<1 mm) circumferential resection margin involvement in esophageal and junctional cancer. Ann Surg Oncol  2013; 20: 2727– 33. Google Scholar CrossRef Search ADS PubMed  24 Sillah K, Pritchard S, Watkins G et al.   The degree of circumferential tumour involvement as a prognostic factor in esophageal cancer. Eur J Cardiothorac Surg  2009; 36: 368– 73. Google Scholar CrossRef Search ADS PubMed  25 Lee G D, Lee S E, Kim K M et al.   New 3-tiered circumferential resection margin criteria in esophageal squamous cell carcinoma. Ann Surg  2015; 262: 965– 71. Google Scholar CrossRef Search ADS PubMed  26 Chao Y, Yeh C, Chang H et al.   Impact of circumferential resection margin distance on locoregional recurrence and survival after chemoradiotherapy in esophageal squamous cell carcinoma. Ann Surg Oncol  2011; 18: 529– 34. Google Scholar CrossRef Search ADS PubMed  27 Harvin J, Lahat G, Correa A et al.   Neoadjuvant chemoradiotherapy followed by surgery for esophageal adenocarcinoma: significance of microscopically positive circumferential radial margins. J Thorac Cardiovasc Surg  2012; 143: 412– 20. Google Scholar CrossRef Search ADS PubMed  © The Authors 2017. Published by Oxford University Press on behalf of International Society for Diseases of the Esophagus. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

Diseases of the EsophagusOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial